ABCDEFG

UNIVERS ITY OF OULU P.O.B . 7500 F I -90014 UNIVERS ITY OF OULU F INLAND

A C T A U N I V E R S I T A T I S O U L U E N S I S

S E R I E S E D I T O R S

SCIENTIAE RERUM NATURALIUM

HUMANIORA

TECHNICA

MEDICA

SCIENTIAE RERUM SOCIALIUM

SCRIPTA ACADEMICA

OECONOMICA

EDITOR IN CHIEF

PUBLICATIONS EDITOR

Senior Assistant Jorma Arhippainen

University Lecturer Santeri Palviainen

Docent Hannu Heusala

Professor Olli Vuolteenaho

University Lecturer Hannu Heikkinen

Director Sinikka Eskelinen

Professor Jari Juga

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-0115-3 (Paperback)ISBN 978-952-62-0116-0 (PDF)ISSN 0355-3191 (Print)ISSN 1796-220X (Online)

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

OULU 2013

A 608

Eeva Jansson

PAST AND PRESENT GENETIC DIVERSITY AND STRUCTURE OF THE FINNISH WOLF POPULATION

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF SCIENCE,DEPARTMENT OF BIOLOGY

A 608

ACTA

Eeva Jansson

A C T A U N I V E R S I T A T I S O U L U E N S I SA S c i e n t i a e R e r u m N a t u r a l i u m 6 0 8

EEVA JANSSON

PAST AND PRESENT GENETIC DIVERSITY AND STRUCTURE OF THE FINNISH WOLF POPULATION

Academic dissertation to be presented with the assent ofthe Doctoral Training Committee of Technology andNatural Sciences of the University of Oulu for publicdefence in Auditorium TS101, Linnanmaa, on 24 May2013, at 12 noon

UNIVERSITY OF OULU, OULU 2013

Copyright © 2013Acta Univ. Oul. A 608, 2013

Supervised byProfessor Jouni AspiDocent Minna RuokonenDoctor Jaana Liimatainen

Reviewed byProfessor Pekka PamiloProfessor Juha Kantanen

ISBN 978-952-62-0115-3 (Paperback)ISBN 978-952-62-0116-0 (PDF)

ISSN 0355-3191 (Printed)ISSN 1796-220X (Online)

Cover DesignRaimo Ahonen

JUVENES PRINTTAMPERE 2013

Jansson, Eeva, Past and present genetic diversity and structure of the Finnish wolfpopulation. University of Oulu Graduate School; University of Oulu, Faculty of Science, Department ofBiology, P.O. Box 3000, FI-90014 University of Oulu, FinlandActa Univ. Oul. A 608, 2013Oulu, Finland

Abstract

Many species and populations have perished as a consequence of human actions. During the last~200 years, large carnivores have been almost completely extirpated from Western Europe. Large-scale wolf hunting started in Finland around the 1850s, and the population size quickly collapsed.The population was very small until the mid-1990s, when wolves started to regularly reproduce inFinland again. The wolf is an endangered species in Finland, and the biggest threat to the species’survival is excessive hunting.

In this doctoral thesis study, I inspected the genetic structure and diversity of the Finnish wolfpopulation using neutral genetic markers. Almost 300 wolves from the contemporary Finnishpopulation and over 50 wolves from the north-western Russia were analyzed with geneticmethods. Additionally, the genetic history of the population was examined with the help of over100 museum samples.

The modern Finnish wolf population proved to be genetically as diverse as the non-endangeredEastern European and North American wolf populations. However, the genetic diversitydecreased significantly during the study period (1995–2009), and was at its lowest level in the finalphase of the examination. In tandem, the inbreeding coefficient rose to a relatively high level.Genetic sub-structures were observed due to social structures within wolf packs. The meandispersal distances of wolves were approximately only 100 km. The Finnish wolf population isdivided into neighbourhoods of related individuals, and their size substantially decreased duringthe study period. This pattern, together with the growth of the inbreeding coefficient, suggests thatlost alpha individuals in wolf packs are replaced by their offspring.

This study demonstrated that Russian and Finnish wolf populations are nowadays geneticallydifferentiated. Gene flow between the populations is low, despite the geographic interconnection.Only a few possible immigrants from Russia into Finland were detected in the study. The effectivesize of the Finnish wolf population proved to be small, and was mainly below the often-consideredcritical size of 50. Historical analysis revealed that the Finnish wolf population was formerlygenetically more diverse, more continuous with the Russian wolf population, and had a more than90% larger effective size.

On the basis of this study, the genetic status of the Finnish wolf population is worrying andneeds to be monitored. The population should be substantially larger than today and/or the amountof gene flow higher, so that the population viability could be considered secured even in the shortterm.

Keywords: Canis lupus, conservation genetics, effective population size, gene flow,inbreeding, neutral genetic variation, population bottleneck

Jansson, Eeva, Suomen susipopulaation entinen ja nykyinen geneettinen rakenneja monimuotoisuus. Oulun yliopiston tutkijakoulu; Oulun yliopisto, Luonnontieteellinen tiedekunta, Biologian laitos,PL 3000, 90014 Oulun yliopistoActa Univ. Oul. A 608, 2013Oulu

Tiivistelmä

Ihmisen toiminnan seurauksena lukuisat eliölajit ja –populaatiot ovat hävinneet. Viimeisten noin200 vuoden aikana suurpedot hävitettiin lähes koko Länsi-Euroopasta. Laajamittainen suden-metsästys alkoi Suomessa 1850-luvun paikkeilla ja kanta romahti nopeasti. Populaatio oli hyvinpieni lähes koko 1900-luvun, ja sudet ovat jälleen lisääntyneet yhtäjaksoisesti Suomessa vasta1990-luvun puolivälistä. Susi on erittäin uhanalainen Suomessa ja merkittävin uhka lajin säily-miselle on liiallinen metsästys.

Tarkastelen tässä väitöskirjatyössäni Suomen susipopulaation geneettistä rakennetta ja moni-muotoisuutta neutraaleja geenimerkkejä käyttäen. Tutkimuksessa analysoitiin geneettisin mene-telmin lähes 300 sutta nyky-Suomesta sekä yli 50 sutta Luoteis-Venäjältä. Lisäksi populaationgeneettistä historiaa selvitettiin yli 100 museonäytteen avulla.

Nykyinen Suomen susipopulaatio osoittautui tutkimuksessa geneettisesti yhtä monimuotoi-seksi kuin ei-uhanalaiset susipopulaatiot Itä-Euroopassa ja Pohjois-Amerikassa. Geneettisenmuuntelun määrä kuitenkin laski tutkimusajanjaksolla (1995–2009) merkitsevästi ollen matalintarkastelujakson lopussa. Samanaikaisesti populaation sukusiitoskerroin nousi verrattain kor-keaksi. Susipopulaatiossa havaittiin sosiaalisista rakenteista johtuvia geneettisiä alarakenteita.Susien dispersaalimatkat olivat keskimäärin vain noin 100 km. Suomen susipopulaatio on jakau-tunut toisilleen sukua olevien yksilöiden naapurustoiksi, joiden koko pieneni huomattavasti tut-kimusajanjaksolla. Tämä yhdessä sukusiitoskertoimen kasvun kanssa viittaa susilaumojen mene-tettyjen alfayksilöiden korvautumiseen jälkeläisillään.

Tutkimus osoitti, että Venäjän ja Suomen susipopulaatiot ovat nykyisin geneettisesti erilaistu-neet. Geenivirta populaatioiden välillä on maantieteellisestä yhteydestä huolimatta vähäistä. Tut-kimuksessa havaittiin vain muutamia todennäköisiä immigrantteja Venäjältä Suomeen. Suomensusipopulaation efektiivinen koko osoittautui pieneksi ollen pääosin alle kriittisenä rajana pide-tyn 50:en. Historiallinen tarkastelu osoitti Suomen susipopulaation olleen aiemmin geneettisestimonimuotoisempi, yhtenäisempi Venäjän susipopulaation kanssa ja efektiiviseltä kooltaan yli90 % nykyistä suurempi.

Tutkimuksen perusteella Suomen susipopulaation geneettinen tila on huolestuttava ja tarvit-see seurantaa. Populaation tulisi olla nykyistä huomattavasti suurempi ja/tai geenivirran määränkorkeampi, jotta populaation elinvoimaisuuden voitaisiin katsoa olevan turvattu edes lyhyelläaikavälillä.

Asiasanat: Canis lupus, efektiivinen populaatiokoko, geenivirta, neutraali geneettinenmuuntelu, populaation pullonkaula, sukusiitos, suojelugenetiikka

In memory of Minna.

Thank you for teaching me so much.

8

9

Acknowledgements

This study was a long journey and I have very many people to thank for their

help, encouragement and support. First of all, I would like to thank my supervisor,

Jouni Aspi, for giving me the opportunity to study wolf genetics, but especially

for sharing his broad expertise and know-how with me and giving the help I

needed over the years. I really enjoyed our discussions, through which I learned a

lot!

I’m most grateful to my second supervisor, Minna Ruokonen, who passed

away shortly before this study was finished. She taught me all the basics of lab

work and always gave very good and helpful feedback. I’m very sad I cannot

share the joy of finishing this study with her, but will try to keep in mind

everything I learned from her, and hopefully continue my scientific career in a

way she would have been proud of.

I am also deeply grateful to my third supervisor, Jaana Liimatainen. Our

weekly lunch meetings gave me incredibly much! With you I could share all my

joys and sorrows of work and life in general, and with the help of your

encouragement I continued when I wanted to give up. By the way, is it my or your

turn to pay the next coffee?

My compliments to Ilpo Kojola and other people from the Finnish Game and

Fisheries Research Institute and Evira, who work with wolves and provide us

with the samples and information we need(ed) for our studies. I also want to

express my gratitude to other co-authors and to Professors Pekka Pamilo and Juha

Kantanen for pre-examination of this thesis.

Big thanks to all current and past group members, with whom I have had the

privilege to work and become friends. Special thanks to Alex, Julia, Alina, Jenni,

Matti, Marja, Rodrigo, Jukka, and Liisa S. I also express my gratitude to the staff

at the Botanical Gardens and Botanical Museum, who gave me the most

welcoming work community when I needed one. Thank you Päivi, Ulla and

Johanna for sharing a room and/or experiences with me. I would also like to thank

you Liisa R. for the time we worked together with flies, and Katja, who was a part

of our coffee team during that time.

My compliments to the Population Genetics Graduate School for funding and

all the interesting courses I could participate in. In particular, I would like to

thank the Chair of the School, Professor Outi Savolainen, who was also the chair

of my follow-up group. Other funding for this project was provided as personal

grants from the Alfred Kordelin Foundation, Ella & Georg Ehrnrooth Foundation,

10

and the Faculty of Science, University of Oulu, to which I hereby want to express

my gratitude.

I would also like to thank technicians Soile and Hannele for their help with

my museum samples as well as Pekka and Kai for all their help in the clean

facilities.

Lastly I thank my friends and family. Ulla, Helena, Riikka and all my other

friends: Thank you for being there for me! My warmest thanks for everything to

my parents Hannu and Tuula and my godmother Anja, my siblings Juha, Leila

and Anu, their respective partners Hanne, Göran and Antti, and to my parents-in-

law Anders and Gunn. Finally, I thank my beloved husband Niklas. You are my

everything.

Tornio, March 2013 Eeva Jansson

11

Abbreviations

A number of alleles

AR allelic richness

b slope

bp base pair

cpDNA chloroplast DNA

DNA deoxyribonucleic acid

FIS departure from Hardy-Weinberg proportions within subpopulations,

local inbreeding coefficient

FST proportion of genetic variation due to differences among

populations, index of population differentiation

H number of haplotypes

Hd haplotype diversity

He expected heterozygosity

Ho observed heterozygosity

HR haplotype richness

IBD isolation by distance

K number of clusters

LD linkage disequilibrium

mtDNA mitochondrial DNA

m proportion of immigrants, migration rate

N number of, size of an ideal population

Nb number of breeding individuals, neighbourhood size

Nc census population size

Ne effective population size

Nm average number of immigrants

P statistic test value

PCR polymerase chain reaction

PR private haplotype richness

S number of polymorphic sites

SD standard deviation

π nucleotide diversity

12

13

List of original articles

This thesis is based on the following articles, which are referred to in the text by

their Roman numerals.

I Aspi J, Roininen E*, Ruokonen M, Kojola I & Vilà C (2006) Genetic diversity, population structure, effective population size and demographic history of the Finnish wolf population. Molecular Ecology 15: 1561–1576.

II Aspi J, Roininen E*, Kiiskilä J, Ruokonen M, Kojola I, Bljudnik L, Danilov P, Heikkinen S & Pulliainen E (2009) Genetic structure of the northwestern Russian wolf populations and gene flow between Russia and Finland. Conservation Genetics 10: 815–826.

III Jansson E, Ruokonen M, Kojola I & Aspi J (2012) Rise and fall of a wolf population: Genetic diversity and structure during recovery, rapid expansion, and drastic decline. Molecular Ecology 21: 5178–5193.

IV Jansson E, Harmoinen J, Ruokonen M & Aspi J (2013) Reconstructing the genetic history of the Finnish wolf population during the last 150 years. Manuscript.

*Jansson née Roininen

14

15

Table of contents

Abstract

Tiivistelmä

Acknowledgements 9 Abbreviations 11 List of original articles 13 Table of contents 15 1 Introduction 17

1.1 The grey wolf (Canis lupus) ................................................................... 17 1.1.1 Distribution ................................................................................... 17 1.1.2 Habits and habitats ....................................................................... 19

1.2 History and current status of the Finnish wolf population ...................... 21 1.3 Population viability ................................................................................. 24

1.3.1 Small population size and the role of genetic factors ................... 24 1.4 Amount of genetic variation .................................................................... 26 1.5 Inbreeding ............................................................................................... 26 1.6 Population bottlenecks ............................................................................ 27 1.7 Population structure ................................................................................ 27 1.8 Population connectivity and gene flow ................................................... 28 1.9 Effective population size ......................................................................... 30 1.10 The use of molecular markers in population genetic studies .................. 32

1.10.1 Microsatellites .............................................................................. 32 1.10.2 Mitochondrial DNA ..................................................................... 33

1.11 Aims of the study .................................................................................... 34 2 Material and methods 35

2.1 Wolf samples ........................................................................................... 35 2.2 DNA extraction, genotyping and sequencing .......................................... 36

2.2.1 Handling of museum samples ...................................................... 37 2.3 Genetic analysis ...................................................................................... 38

2.3.1 Amount of genetic variation ......................................................... 38 2.3.2 Inbreeding ..................................................................................... 38 2.3.3 Population bottleneck tests ........................................................... 39 2.3.4 Population structure ...................................................................... 40 2.3.5 Population connectivity and gene flow ........................................ 41 2.3.6 Effective population size .............................................................. 43

16

3 Results and discussion 45 3.1 Amount of genetic variation .................................................................... 45 3.2 Inbreeding ............................................................................................... 48 3.3 Population bottlenecks ............................................................................ 49 3.4 Population structure ................................................................................ 51

3.4.1 Cryptic population structure – clustering analyses ....................... 51 3.4.2 Limited dispersal – IBD ............................................................... 52

3.5 Population connectivity and gene flow ................................................... 54 3.5.1 Genetic differentiation between Finnish and Russian

wolves ........................................................................................... 54 3.5.2 Amount of gene flow and detection of immigrants ...................... 55

3.6 Effective population size ......................................................................... 57 4 Conclusions 61 References 65 Original articles 77

17

1 Introduction

The direct and indirect influence of human activities has significantly elevated the

extinction rate of species during the past few centuries, and as a consequence

many species and populations are nowadays endangered. This process, commonly

referred as the sixth mass extinction (e.g. Barnosky et al. 2011), causes a loss of

biological diversity, and is especially striking in large mammals (Cardillo et al.

2005) and other apex consumers due to the large cascade effects of these species

on many biological processes and ecosystems (Estes et al. 2011). For instance,

recent extinctions of many carnivore populations in Europe and South and North

America have caused taxonomically depleted guilds, and many of the remaining

populations are likely to be too small to function effectively in the surrounding

ecosystems (Dalerum et al. 2009). Scientific knowledge of the threatened

populations can – if and when adopted in practice – provide us with the necessary

means to manage the remaining populations so that they will be viable in the

future.

1.1 The grey wolf (Canis lupus)

The grey wolf (Canis lupus Linnaeus, 1785; hereafter ‘wolf’) is a charismatic and

controversial species, and one of the most researched wild animals (Mech &

Boitani 2003; Mech 2012). Since the 1990s, molecular markers have widely been

adopted to answer important questions of the species’ biology, history and

evolution, and numerous molecular studies concerning wolves have already been

published (among the first studies, e.g., Kennedy et al. 1991; Lehman et al. 1992;

Roy et al. 1994; Ellegren et al. 1996).

1.1.1 Distribution

The wolf is the largest wild species today belonging to the Canidae family and is

native to Eurasia, North America and North Africa. In recent historical times, the

wolf had the largest distribution area of all terrestrial mammals (Fig. 1; Boitani

2000).

18

Fig. 1. World-wide distribution area of the grey wolf. The distribution in 2003

(“Present”) is shown in the darkest grey, and lighter grey areas indicate the historical

distribution range. (The map is freely available at Wikimedia commons, from where it

was downloaded on 27 August 2012. The original colours were converted to

greyscale.)

In Western and Central Europe, the wolf has been a part of the carnivore

community for at least ~500 000 years, and its direct precursors, Canis

mosbachensis and C. etruscus, already existed about 1 and 1.8 million years ago,

respectively (Croitor & Brugal 2010). Wolves spread to the northern parts of

Europe after the last glaciation and the retreat of the ice sheet approximately 10

000 years ago.

Today, wolf populations are commonly restricted to the wilderness and

remote areas (Mech & Boitani 2004). The largest continuous distribution areas

are found in North America and Asia (Fig. 1; Boitani 2003). Most wolf

populations in Western Europe are nowadays rather small and isolated (Randi

2011). The largest population in Western Europe is located in the Iberian

Peninsula, comprising about 2500 wolves (IUCN 2012), but with no connection

to other populations (Sastre et al. 2011). However, Eastern European wolf

populations (Dinaric-Balkan, Carpathian and Baltic) are larger and more

connected to each other and via Russia to the Asian distribution area (Mech &

Boitani 2004; Linnell et al. 2008). Currently, wolves in Europe might number

approximately 16 000–20 000 individuals (including the European part of Russia;

Linnell et al. 2008).

Wolves were abundant in Europe until the early 19th century (Wayne et al.

1991). The disappearance of wolves from many areas and the extensive world-

19

wide population fragmentation during the last ~200 years coincided with the

expansion of human populations, as the encroachment of human-occupied land

has reduced the number of suitable habitats for wildlife. Besides clear cuttings of

large forest areas, habitat fragmentation and the consequent shortage of available

prey (Fernández & Ruiz de Azua 2010), western wolf populations have vanished

due to active persecution. The other three large predators native to Europe, the

bear, lynx and wolverine, have also disappeared from many western countries

(e.g. Enserink & Vogel 2006). Despite extensive predator removal programmes

that were implemented from the 19th century onwards, including in Eastern

Europe (Pulliainen 1980; Flagstad et al. 2003), substantially larger wolf

populations survived in eastern European and Balkan countries, and many of

them are thriving today (Boitani 2000; Mech & Boitani 2004; Linnell et al. 2008).

The main motive for wolf control was – and still is – economic conflicts with

humans (e.g. Woodroffe 2000; Boitani 2000; Bisi & Kurki 2008; Bisi et al. 2010).

Wolves prey on a range of game species and livestock (Boitani 2003), and have

often been seen as enemies to be feared, persecuted and extirpated (e.g. Mech &

Boitani 2004; Bisi & Kurki 2008). The most negative attitudes towards wolf

presence are typically among hunters, and people living in areas where wolves

have been absent for a long time (e.g. Ericsson & Heberlein 2003; Bisi et al.

2010). Wolf fear and hatred are further increased with prejudices, legends and

misinterpretations of its biology (Boitani 2000).

During recent decades, large carnivores have been making a comeback to

Western Europe, and expanding their territories into areas where they have been

absent for long periods of time (Lucchini et al. 2002; Enserink & Vogel 2006;

Linnell et al. 2008). This increase is mainly due to changes in legislation: Wolves

are nowadays protected in most European countries (Linnell et al. 2008; Randi

2011) and have naturally spread from the remaining populations as hunting

pressure has been reduced. However, wolf conservation and management remain

problematic because of widespread illegal or incidental killing (Caniglia et al.

2010; Randi 2011; Liberg et al. 2012), and the fact that many present-day wolf

populations are still too small to be viable in the long term (see 1.3).

1.1.2 Habits and habitats

Wolves are ecologically flexible (Mech & Boitani 2003) and have a high dispersal

potential of up to several hundreds of kilometres (Wabakken et al. 2001; Valière

et al. 2003, Seddon et al. 2006; Ciucci et al. 2009), which explains their large

20

distribution area. Wolves mostly eat large ungulates such as moose, deer, elk, and

wild boar, but the source of nourishment may vary opportunistically from any

smaller prey item to livestock, carcasses or even garbage (Mech & Boitani 2004).

Occasionally, wolves may also feed on fruits or other plant material (Mech &

Boitani 2003). In fact, wolves can survive anywhere having enough of something

suitable to eat, and where they are not killed by humans (Boitani 2000). As top

predators, wolves have a large impact on ecosystems with effects on prey species

and vegetation (Smith et al. 2003; Hebblewhite et al. 2005; Sergio et al. 2008;

Estes et al. 2011; Ripple & Beschta 2012).

Except for their varying diet composition, wolves are highly adaptable to

various environments and occupy a wide range of habitats from warm deserts to

cold tundra (Wayne et al. 2004). As a part of such adaptation, wolves display

substantial phenotypic variation in body size, weight and colouration, for instance

(e.g. Weckworth et al. 2005; Musiani et al. 2007), and several subspecies of

Canis lupus have been described (Boitani 2000; Mech & Boitani 2004).

Social behaviour is a common feature amongst all modern larger

representatives of Canidae (Croitor & Brugal 2010). Strong social bonds and a

strict hierarchy are peculiar to wolf packs. Wolves are social animals that live

within their territories in packs commonly consisting of a family with one

breeding pair (alpha wolves) and their offspring from one or more litters (Mech &

Boitani 2003). Thus all members of the pack are usually related, apart from the

alpha pair. Pack members co-operate in hunting, reproducing and territorial

defence (Boitani 2000). Typically, a wolf pack consists of 2–15 individuals, and 1

to 8 cubs are born to a breeding pair annually. In Finland the mean litter size is

4.3 ± 1.3 pups (Kaartinen et al. 2010). Large packs are extremely rare in Europe

as a consequence of human-mediated control (Boitani 2000). In North-America,

however, much larger (up to 42 members; Fuller et al. 2003) and more socially

diverse packs (e.g. with subordinate breeders: vonHoldt et al. 2008; Stenglein et

al. 2011) have been found. The natural kin-based social organization of wolf

packs may be interrupted by disturbance (Brainerd et al. 2008). For example,

intense harvesting of wolves in established packs has been found to increase the

adoption of unrelated individuals (Grewal et al. 2004; Jędrzejewski et al. 2005;

Rutledge et al. 2010).

Wolves require large territories and home ranges, and population densities

therefore remain rather low, even in saturated habitats. In Finland, the mean

territorial size is 1372 ± 514 km2 (Kaartinen et al. 2005). Home range sizes vary

considerably and depend on habitat characteristics, prey availability and

21

population density, among other factors (Fuller et al. 2003). Occupied territories

are maintained through howling, scent-marking and direct killing (Mech 1974). In

Finland, wolves leave their natal pack typically as yearlings around the time when

new cubs are born (from March to May; Kojola et al. 2006), in order to find their

own mates and territories (Mech & Boitani 2003). Compared with the high

dispersal capability of wolves, typical dispersal distances have often been

detected to be rather modest (e.g. vonHoldt et al. 2008; Scandura et al. 2011), and

in Finland they are around 100 kilometres (based on radio and GPS transmitters;

Kojola et al. 2006). Short mean dispersal distances and the high stability of

already established wolf packs may lead to local genetic differentiation (i.e.

cryptic population structure; e.g. Scandura et al. 2011; Randi 2011). In general,

demographically stable wolf populations seem to be characterized by very limited

dispersal and gene flow, whereas long-range dispersal is more common during

population expansion and re-colonization phases (Randi 2011).

1.2 History and current status of the Finnish wolf population

The Finnish wolf population represents the most north-western edge of a large,

Eurasian distribution area (Boitani 2003; Fig. 1). The Finnish wolf population has

generally been considered to be continuous with the Russian Karelian population

(Boitani 2003; Linnell et al. 2008). Indeed, the Finnish wolf population did

follow the abundance trends in the Karelian population (Pulliainen 1965; 1980)

until quite recently (Kojola & Määttä 2004).

At the beginning of the 19th century, wolves inhabited the whole area of

Finland (Pulliainen 1980). As in other countries in Western Europe, active hunting

to exterminate wolves from Finland started around the 1850s (Ermala 2003; Bisi

et al. 2010), and led to a rapid population decline. At first, approximately 300–

400 wolves were killed annually in organized hunts (Ermala 2003; Fig. 2), and by

the beginning of the 20th century wolves were only present in eastern and northern

Finland (Pulliainen 1980; Boitani 2003; Ermala 2003), and as a result, the number

of hunted wolves heavily declined (Fig. 2).

22

Fig. 2. The number of recorded wolf kills in Finland in 1845–2010 (redrawn from

Ermala 2003, data for 1991–2010 acquired from the Finnish Game and Fisheries

Research Institute and added to the graph). The figure is included in paper (IV).

It is not known with certainty whether the Finnish wolf population ever vanished

altogether, but nevertheless it remained very small for a long time, and

presumably experienced severe demographic bottlenecks at least during the 1920s

and 1970s (Pulliainen 1965; 1980). It has been argued that in the 1920s the

population even went extinct due to an outbreak of distemper (Mäensyrjä 1974).

The wolf became protected in Finland outside the reindeer husbandry area in

1973 (Anonymous 2005), but until 1995 it was listed as a normal game species,

and the population was controlled by hunting (Bisi et al. 2007). After the

accession of Finland into the European Union, the legislation concerning the

conservation status of the wolf was tightened, and according to the EC Habitats

Directive, the wolf is now strictly protected (Annex IV), except within the

reindeer herding area, where hunting is possible (Annex V). These changes in

legislation probably contributed to the population growth (Anonymous 1996) and

wolves have regularly reproduced again in Finland from the mid-1990s onwards

(Kojola & Määttä 2004; Fig. 3).

23

Fig. 3. The minimum population size (Nc) and number of breeding pairs (Nb/2) in the

Finnish wolf population during 1997–2010. The number of shooting licenses and

recorded annual deaths are also shown and follow the same scale as Nc. The figure

was originally published in paper (III).

After the initial recovery, the Finnish wolf population grew rapidly (Fig. 3) and

expanded into new areas (Kojola et al. 2006; 2011). During the peak year of

2006, 25 breeding pairs were detected and the minimum population size estimate

was ~250 individuals (Fig. 3). However, after 2006, the trend turned to the

opposite direction and the population quickly decreased. This is striking, because

the wolf is listed as an endangered species in Finland (Rassi et al. 2010) and a

special shooting license is always required for hunting. The annual legal harvest

is at most ~15% of the population census size (Kojola et al. 2011; Fig. 3), and no

disease epidemics – that could also have explained the heavy decline – have been

reported. This strongly suggests that illegal killing might explain the recent

population decline (Kojola et al. 2011), but no formal studies on this subject have

been conducted in Finland (but see Liberg et al. 2012 for a similar case study on

the neighbouring Scandinavian population). According to the latest estimate from

24

the Finnish Game and Fisheries Research Institute (February 2013), there are now

only 120–135 wolves in Finland.

1.3 Population viability

Large carnivores are important ecosystem components, but prone to extinction

due to often small and fragmented populations, slow growth rates and large area

requirements (Cardillo et al. 2005). Additionally, carnivores have suffered a high

level of human persecution, both historically and recently (Woodroffe 2000;

Dalerum et al. 2009). Although the wolf as a species worldwide is considered to

be of least concern (IUCN 2012), and has a high reproductive potential compared

to many other large carnivores, wolf populations in many European countries are

considered to be well below the threshold of a viable population (Boitani 2000;

see also Traill et al. 2010 and Flather et al. 2011).

Population viability is a complex concept comprising two interacting

components: the demographic and the genetic. Demographic models often

determine viability as the probability of extinction within a certain number of

years, and therefore set levels on the minimum viable population size (MVP) (e.g.

Boyce 1992). The genetic viability concept includes, besides short-term survival,

a long-term evolutionary perspective: In the short time frame, genetic diversity

provides a buffer against environmental fluctuations, whereas in the long term,

genetic diversity provides the raw material for natural selection and the ability to

adapt to the changing environment (e.g. Frankham 2005; Laikre et al. 2009).

Knowledge of the genetic properties of populations of conservation concern, such

as effective sizes and connectivity with other populations, is therefore of

considerable relevance in management and should not be omitted, for instance, in

assessment of the favourable conservation status of populations (Laikre et al.

2009). Genetic information may also provide insights into the demographic

structure and history of a population, which are valuable for longer-term

conservation planning (e.g. Carmichael et al. 2007; Yang & Jiang 2011;

Thalmann et al. 2011).

1.3.1 Small population size and the role of genetic factors

The study of small population size represents one of the foundations of

conservation biology (Bouzat 2010). This is because population size is the major

25

determinant of population well-being and extinction risk (Reed et al. 2003), and

small populations are in many ways more prone to genetic erosion.

Extinction is a demographic process caused by deterministic and/or stochastic

factors. Deterministic factors, such as habitat destruction and overexploitation

inevitably threaten populations when realized, whereas stochastic threats are

random, unpredictable changes in environmental, demographic and genetic

factors (Allendorf & Luikart 2007, pp. 10–11). Despite their randomness, the

abundance and/or the impact of stochasticity is somewhat bound to population

size. Stochasticity will become elevated in small populations and may lower the

population fitness and increase the risk of extinction (Frankham 2005; Allendorf

& Luikart 2007; Palstra & Ruzzante 2008). Genetic stochasticity includes random

genetic change via drift (see 1.4) and increased inbreeding (see 1.5), which leads

to the loss of genetic variation and an increase in the frequency of harmful allelic

combinations (Allendorf & Luikart 2007, p. 10). It is noteworthy that the strength

of natural selection is also conditional: natural selection becomes less effective in

small populations, and may be swamped by genetic drift as a major evolutionary

force (e.g. Allendorf & Luikart 2007, p. 186). If the selection coefficient, s, is

smaller than 1/Ne for an allele, it will act as if it were selectively neutral (Li

1978). Thus, slightly deleterious mutations can accumulate over time in small

populations and slightly beneficial alleles can be lost.

Small population size may also increase the risk of hybridization when the

probability of finding a mate of the same species is limited (e.g. Allendorf &

Luikart 2007, p. 428). Several cases have been reported of introgression from

dogs (e.g. Vilà et al. 2003b; Muñoz-Fuentes et al. 2010; Hindrikson et al. 2012)

or other canids (e.g. coyotes; Fain et al. 2010; Rutledge et al. 2011) to wolf

population gene pools. In conservation, interspecific hybridization is generally

thought to be harmful because it can result in the disappearance of a distinct taxon

and/or loss of specific adaptations (e.g. Muñoz-Fuentes et al. 2010). Additionally,

hybrids may behave abnormally, which can complicate the protection of the

endangered species in question (Fain et al. 2010).

In the following chapters, some of the central population genetic concepts

related to conservation and small population size are presented. The study of

genetic variation (1.4), inbreeding (1.5), population bottlenecks (1.6) and internal

structure (1.7), together with population connectivity and gene flow (1.8) and

effective population size (1.9) form the body of this work, and all or most of the

subjects are covered in papers I–IV.

26

1.4 Amount of genetic variation

The amount of genetic variation in populations is dictated by the interplay of local

population size, gene flow, natural selection, and ultimately the underlying

mutation rate (e.g. Frankham 1996). The mutation rate in general is very low, and

in practice its role in a short time frame is very small. Therefore, population size

(or more precisely, effective population size; see 1.9) and gene flow (1.8) have a

more central role, for instance, in conservation and population management.

Genetic diversity is acknowledged as a central part of biological diversity and

important for population viability in the short and long term (Laikre et al. 2009).

Natural populations are finite in size, which restricts the amount of genetic

diversity (i.e. the number of alleles, heterozygosity level and polymorphism) in a

population at a certain time. Additionally, genetic drift causes random changes in

allele frequencies between generations due to sampling error (Allendorf &

Luikart 2007, p. 118), and neutral genetic variation is expected to be lost at a rate

inversely proportional to effective population size (e.g. Frankham 1996). On the

contrary, gene flow among populations is efficient in counteracting the loss of

genetic variation (see 1.8), and in the long term genetic diversity is best preserved

in large populations and/or in populations characterized by recurrent gene flow.

1.5 Inbreeding

Inbreeding is often considered one of the major genetic threats in small

populations (e.g. Hedrick & Kalinowski 2000; Frankham 2005; O’Grady et al.

2006), where mating between related individuals is inevitable over time. The

effect of heterozygosity loss on fitness reduction is often slow and associated with

environmental stress, whereas the growth of inbreeding and the accompanying

inbreeding depression have an immediate impact (Frankham 2005). Inbreeding

depression refers to the lower fitness of inbred individuals due to shared ancestry,

and can be explained with two genetic mechanisms: an increase in homozygosity

and, by implication, the expression of recessive deleterious alleles, and/or a

reduced frequency of heterozygote genotypes with better fitness (e.g. Keller &

Waller 2002; Allendorf & Luikart 2007, pp. 307–308).

Inbreeding depression is often very difficult to detect in nature (e.g. Allendorf

& Luikart 2007), but its negative effects have been confirmed in wild wolves as

decreased over-winter survival of inbred pups (in Scandinavia; Liberg et al.

2005). Increased congenital bone deformities have also been detected in inbred

27

and isolated wolf populations (Scandinavia, Isle Royale; Räikkönen et al. 2006,

2009). In a captive wolf population, one form of recessive hereditary blindness

was confirmed to be expressed frequently due to inbreeding (Laikre et al. 1993).

Usually, wolves strictly avoid mating with close relatives (e.g. Smith et al.

1997; vonHoldt et al. 2008; Rutledge et al. 2010), but in cases where a breeding

individual is lost close to the mating season and no unrelated mate is available,

close inbreeding might occur (Liberg et al. 2005). A recent study by Geffen et al.

(2011) demonstrated that the rate of inbreeding in social canids might be related

to the proximity of close relatives and that pack members are excluded as mates,

but that mating outside natal packs would be random (i.e. indiscriminate with

respect to relatedness). Thus, the active avoidance of inbreeding in wolves can be

somewhat questionable.

1.6 Population bottlenecks

When a population undergoes a bottleneck, i.e. a significant reduction in its

effective size, it loses its genetic variation due to random drift, when only a small

fraction of individuals survive and reproduce in the following generations.

Inbreeding is also likely to increase as a consequence of a bottleneck (see above).

Equivalent situations to a population bottleneck are the founding or re-

establishment of a population from a larger source population; only a proportion

of individuals contribute to the new gene pool (Allendorf & Luikart 2007, p. 127).

The genetic outcome of demographic bottlenecks may not always be

straightforward, however, and chance, selection and population history can play

large roles (Bouzat 2010).

A major population size reduction often leaves signals in the genetic pool of a

population, and may therefore be subsequently detectable by analysing post-

bottleneck patterns of genetic variability. The principles of the statistical testing of

population bottlenecks are discussed in detail in chapter 2.3.3.

1.7 Population structure

The genetic structure of any population is a complex interplay of past historical

events, species behavioural traits, geographical features limiting gene flow and

complex ecological processes shaping it (Pilot et al. 2006). Among historical

factors, the last glaciation is known to have had profound effects on the genetic

architecture of extant species via postglacial re-colonization dynamics (e.g.

28

Hewitt 2000), which in wolves can be detected by the presence of two

geographically distinct mitochondrial haplotype groups between south-western

and eastern Europe (Pilot et al. 2010). More recent historical events and

especially the large anthropogenic influence during the last few centuries have

caused major shifts in the demography, distribution and genetic structures of

many populations (e.g. in wolves, Flagstad et al. 2003; in arctic foxes, Nyström et

al. 2006; in western gorillas, Thalmann et al. 2011). Knowledge of historical

population structures prior to large anthropogenic influences is important, given

that it can be used as a baseline for today’s management goals. In endangered

populations, connectivity among the remaining and often very fragmented

populations is of special interest, because of the homogenizing effect of gene flow

(see 1.8).

Besides historical changes and connectivity to nearby populations, the

presence of cryptic (i.e. hidden) substructures within populations must be taken

into consideration. Such structures are often linked to behavioural traits and

species ecology including philopatry, dispersal capability, sociality, mating

models and food preferences (e.g. vonHoldt et al. 2010; Scandura et al. 2011;

Edelaar & Bolnick 2012; Pilot et al. 2012). Different anthropogenic factors may

also largely affect the demography and spatial dynamics of wild populations (e.g.

Stronen et al. 2012).

1.8 Population connectivity and gene flow

The study of population connectivity (demographic cohesion) and gene flow is

one of the key issues when defining the current status and evolutionary potential

of a population (Waples & Gaggiotti 2006; Palstra & Ruzzante 2008; Lowe &

Allendorf 2010). Natural populations are distributed over geographical space with

varying degrees of connectivity between regions or subpopulations, ranging from

total isolation to panmixia.

Genetic connectivity primarily depends on the absolute number of dispersers

among populations, and can be defined as the degree to which gene flow affects

evolutionary processes within subpopulations (Lowe & Allendorf 2010).

Therefore, the evolutionary population concept emphasizes reproductive

interactions between individuals (Waples & Gaggiotti 2006), i.e. the genetic

elements of the population processes occur at larger spatial and temporal scales

than demographic ones. Consequently, maintaining the evolutionary potential of a

population is a long-term conservation issue that requires much larger numbers of

29

individuals than the short-term maintenance of local populations to avoid

demographic extinction (e.g. Linnell et al. 2008; Laikre et al. 2009; Hansen et al.

2011).

Gene flow, even at very low levels seems to be efficient at countering

inbreeding and its harmful effects (e.g. Vilà et al. 2003a; Adams et al. 2011), and

may maintain the genetic cohesiveness of species by providing selectively

advantageous alleles (Rieseberg & Burke 2001). Additionally, gene flow will

override the effects of random genetic drift when the migration rate, m > 1/4 Ne

(Wright 1931).

It is often difficult to define population limits. Even in situations in which the

movement of individuals across regions or assumed/known subpopulations is

closely monitored (like the movement of wolves across the border between

Finland and Russia; Fig. 4 below) and frequent, the degree to which this

movement translates into genetically effective dispersal is often not well

established (e.g. vonHoldt et al. 2010).

Fig. 4. Number of wolves crossing the Finnish–Russian border counted by border

guards in 1995–2010. Redrawn from Pulliainen 2011.

30

Different genetic methods can be used to determine the degree of reproductive

cohesion between subpopulations. In the absence of true panmixia, a population

will be genetically subdivided and allele frequencies will differ significantly (if

enough loci and individuals are sampled; Waples & Gaggiotti 2006). These

existing genetic differences between (sub-)populations and individuals within

them provide the basis to test the origin of individuals and the amount of gene

flow. The methods used in this study for estimating gene flow and population

differentiation are described in chapter 2.3.5.

As wolves are highly mobile, high rates of gene flow that reduce genetic

differentiation among local populations could be expected (Pilot et al. 2006;

Randi 2011). Dispersal patterns and distances are, however, affected by many

variable factors such as population densities, food availability, social structures,

anthropogenic influence and individual preferences (e.g. Pilot et al. 2006; Croteau

2010; Edelaar & Bolnick 2012). Several recent studies have discovered distinct,

larger-scale hierarchical population units within grey wolves that correspond with

geographic and ecological differences among populations (Carmichael et al.

2001, 2007; Geffen et al. 2004; Weckworth et al. 2005; Musiani et al. 2007;

Muñoz-Fuentes et al. 2009), so dispersing wolves tend to stay close to their natal

area or choose habitats resembling them (e.g. Pilot et al. 2012). Several recent

genetic studies have shown that even on a local scale, wolf populations might be

genetically structured and the amount of gene flow restricted (vonHoldt et al.

2010, Scandura et al. 2011; Stronen et al. 2012). A number of reasons exist for

such short-distance genetic subdivision, among them human-caused population

fragmentation (Stronen et al. 2012) and other anthropogenic factors (vonHoldt et

al. 2010) together with the high spatial stability of existing packs (Scandura et al.

2011).

1.9 Effective population size

The concept of effective population size, Ne, introduced for the first time in 1931

by Sewall Wright, is one of the most fundamental parameters in evolutionary

genetics and conservation biology, because it influences the rate of inbreeding

and loss of genetic variation (Waples 2002). In interaction with systematic forces,

such as natural selection, mutation and migration, the effective size determines

the amount and distribution of genetic variation present in a population (Wang

2005).

31

Random genetic processes in a population occur at rate inversely related its

effective size, which is the size of an 'ideal' population (Wright 1931), with e.g.

random mating, an equal sex ratio, non-overlapping generations and a constant

population size (Waples 2005). Populations with small effective sizes are greatly

affected by genetic drift, and the loss of adaptive potential, and are more prone to

inbreeding depression (Keller & Waller 2002; Frankham 2005). Besides directly

affecting the rate of genetic change, Ne also determines the relative importance of

migration and selection; these forces are deterministic in large populations but

can be overwhelmed by random processes in small ones (Waples 2005).

Therefore, knowledge of Ne and the different factors affecting it besides the

population census size is important in wildlife management and conservation

(Luikart et al. 2010).

As natural populations deviate from assumptions of the ideal population in

many ways, effective sizes are often much smaller than census sizes (Palstra &

Ruzzante 2008). Consequently, Ne/Nc ratios are often relatively low (reported

medians: ~0.5, Nunney & Elam 1994; 0.11, Frankham 1995; and 0.16, Palstra &

Ruzzante 2008) and may also be unstable if the important factors influencing Ne

are not stable. In particular, population size fluctuation, an unequal sex ratio and

variance in reproductive success are expected to have a considerable effect on Ne

(Frankham 1995). Accurate estimation of Ne with demographic methods (see

Allendorf & Luikart 2007, pp. 151–158) requires extensive, often hard-to-gather

knowledge on the life history parameters and reproductive biology of a study

population. They are not applicable, for instance, for elusive or nocturnal species

or for the study of historical populations. Additionally, demographic methods may

often overestimate the true Ne, because they seldom include all important factors,

such as variance in reproductive success, which can reduce the effective size

compared to the census size (Luikart et al. 2010).

Several genetic methods have been developed to estimate Ne (for reviews, see

e.g. Leberg 2005; Wang 2005; Palstra & Ruzzante 2008; Luikart et al. 2010),

which apply different measures of genetic change, and extend to different time

frames and on different geographic scales. The most widely used and evaluated

concepts of Ne are inbreeding (NeI) and variance (NeV), which are concerned with

the rate of loss of heterozygosity (or the increase of homozygosity) and the

change in allele frequencies through time (i.e. drift), respectively (Wang 2005;

Luikart et al. 2010). The effective population size and its changes in the Finnish

wolf population were estimated throughout this study, and the principles of Ne

estimation with genetic methods are presented in 2.3.6.

32

1.10 The use of molecular markers in population genetic studies

Molecular markers are used to study genetic factors and their manifestation in

nature. The rapid development of various techniques and methods to extract,

amplify and analyse genetic variation in organisms since the late 1960’s (for the

first genetic studies see e.g. Lewontin & Hubby 1966; Harris 1966; Selander &

Yang 1969) has in many ways revolutionized our knowledge of genetics and

biology in general. All molecular markers can be divided into three conceptually

different classes: protein variants (allozymes), DNA sequence polymorphisms and

DNA repeat variations (Schlötterer 2004).

Genetic variation can also be divided according to whether it influences

individual fitness into neutral and non-neutral variation. Neutral genetic markers

are not subject to natural selection, and therefore do not intrinsically provide

information about adaptive genetic variation, unless Ne is very small, when the

heritability of quantitative traits is also likely to respectively decrease (Willi et al.

2006, see e.g. Marsden et al. 2012 for an example, where neutral and adaptive

diversity were correlated in an endangered canine population). Neutrally evolving

genetic markers, on the other hand, can be used for instance, to decipher the

demographic changes and genetic structures of populations (e.g. Flagstad et al.

2003; Marsden et al. 2012), in individual identification and family relationship

studies (e.g. Jędrzejewski et al. 2005; vonHoldt et al. 2008) and in phylogenetic

studies (e.g. Pilot et al. 2010).

Wolves and dogs have the same genetic make-up, consisting of 78

chromosomes (38 autosomes and two sex chromosomes). Because the divergence

of dogs from wolves, i.e. the domestication process, is in evolutionary terms very

recent (≤ 16 300 years ago; Pang et al. 2009; but see Druzhkova et al. 2013), the

genomes of these two species have not greatly diverged, and the large number of

polymorphic genetic markers developed for dogs are also directly usable for

wolves.

1.10.1 Microsatellites

Microsatellites are relatively short (~75–300bp; Allendorf & Luikart 2007) DNA

strands with a tandemly repeated sequence motif of one to six base pair(s) in

length (Schlötterer 2000). Since their discovery in the early 1980s, they have

become widely used markers in molecular and wildlife genetic studies due to their

abundance, high polymorphism and biparental inheritance (Ellegren 2004).

33

Microsatellites serve as a versatile tool, for instance, to identify individuals

(‘genetic fingerprinting’), to study family relationships, and to assign and cluster

individuals according to the population/group of origin. A large body of software

has also been developed for microsatellite markers to decipher, for example, the

underlying population structures, demographic histories, the amount of gene flow

between populations, and to reveal possibly important units for conservation (e.g.

Wayne & Morin 2004). Because of their relatively small size, microsatellite

sequences are often also applicable to samples with low DNA quality and/or

quantity, e.g. those obtained from non-invasive sampling schemes (e.g. Stenglein

et al. 2011; Kopatz et al. 2012).

Microsatellites are rather evenly distributed in organisms’ genomes and

mostly selectively neutral (but see Li et al. 2002). Because of their neutral nature,

the degree of variability of microsatellites depends on the underlying mutation

rate (Ellegren 2004). The mutation rate is generally high, ~10−2 − 10−6 events per

locus per generation (Li et al. 2002), and the majority of the changes are single

alterations of the repeat number (i.e. follow the stepwise mutation model, SMM)

rather than variations in the primary sequence (Li et al. 2002; Ellegren 2004). The

main mutational mechanisms of microsatellites are replication slippage and

unequal recombination. In this case, highly repeated genome parts are somewhat

inaccurately copied and aligned during cell division, and some of these mutations

escape the following repair mechanisms (Li et al. 2002).

1.10.2 Mitochondrial DNA

In the first studies concerning the genetic variation in natural populations,

mitochondrial DNA (mtDNA) was investigated (e.g. Avise et al. 1979).

Mitochondria are energy-producing cell organelles containing their own double-

stranded circular DNA, the vast majority of which is inherited maternally (e.g.

Allendorf & Luikart 2007). There is a considerable amount of mtDNA in all

eukaryotic cells. Due to this abundance and the relatively small size compared to

the autosomal genome, mitochondrial DNA is often much better preserved in

specimens of poorer quality. Therefore, mtDNA is widely employed in such

genetic applications as ancient and museum DNA analysis (e.g. Ramakrishnan &

Hadly 2009) and forensics.

Mitochondrial DNA sequences are also suitable for reconstructing

phylogenetic trees (i.e. maternal lineages) since there is generally no

recombination between mtDNA molecules (Allendorf & Luikart 2007; Galtier et

34

al. 2009), and phylogenetic relationships are therefore easy to interpret. A part of

the mtDNA commonly used in genetic analyses is known as the control region,

which does not contain genes and has the highest variability. In dogs and wolves,

the total length of the control region is ~1270 bp (Kim et al. 1998).

Mitochondrial DNA is highly variable in natural populations because of its

high mutation rate, which can generate signals of the population history over

relatively short time frames. Because of the lack of recombination, the whole

mitochondrial genome behaves like a single genetic locus (Galtier et al. 2009),

and its usefulness in many population genetic studies is thus rather limited. In

genetic studies concerning wolves, mtDNA sequences have recently been utilized,

for example, to conclude on the phylogeographic history of wolves in Europe

from ancient samples (Pilot et al. 2010), and to reveal the spatial genetic structure

due to differing ecological factors (Pilot et al. 2006).

1.11 Aims of the study

This thesis presents the first larger-scale genetic study of the Finnish wolf

population. Prior to this study, some Finnish wolf samples had been genetically

analysed only for reference purposes (Vilà et al. 1999; Randi et al. 2000; Flagstad

et al. 2003; Vilà et al. 2003a; Lucchini et al. 2004; Seddon et al. 2006). The aim

of this study was to provide a good general insight into the genetic structure and

diversity of the present-day Finnish wolf population based on neutral genetic

markers, and to determine the current level of genetic cohesion with the

neighbouring north-western Russian population. Important questions to be

answered were the following:

1. What is the level of genetic diversity and how is it distributed in the Finnish

wolf population?

2. What is the inbreeding coefficient and effective population size?

3. Are any substructures or genetic signals of recent bottleneck(s) detectable?

4. How close is the genetic connection with the Russian population (i.e. what

are the levels of gene flow and genetic differentiation between these

populations)?

The amount and distribution of genetic variation in the Finnish wolf population

during historical times were also investigated, and compared to those seen in the

present-day population in order to reveal, whether and how the population has

changed during a period of over 150 years of active hunting.

35

2 Material and methods

The material and methods are only briefly described here. Detailed information

can be found in the original papers (I–IV). Background information underlying

the genetic analyses is given in more detail.

2.1 Wolf samples

Systematic field studies and the collection of wolf samples were started in Finland

in the mid-1990s by the Finnish Game and Fisheries Research Institute, through

which the present-day samples used in this study were obtained. In total, 298

contemporary Finnish wolf samples were analysed (all or part of them used in

papers I–IV). Figure 5 below shows the geographic location for all collected

Finnish samples. In addition, samples from neighbouring Russian Karelian

(N = 39) and Arkhangelsk Oblast (N = 14) were used (II, III).

Fig. 5. Map showing the geographic location of wolf samples analysed during this

study. Black circles correspond to samples from the contemporary Finnish

population, grey circles to historical Finnish samples and white circles show

approximate collection sites for the contemporary Russian wolf samples.

36

Sources of DNA for this research included tissue samples (muscle or liver

collected after the death of an individual), pelt samples or blood collected on

snow, as well as mouth swabs on FTATM cards (Whatman) or hair samples from

living individuals. In addition to the obtained genotypes (see 2.2 below) and data

on the location of the samples, other information gathered by field workers and

from autopsy reports was applied. This included knowledge of family relations,

sex and the reproductive status of the individuals, amongst others. In papers I and

IV the collection date was used as a basis of sample division into approximate

temporal groups. In paper III, samples were divided into five temporal birth

cohorts for the detection of short-term temporal genetic variation. In order to do

so, an estimate for the birth year of each individual was needed. Estimation was

based on field observations and/or tooth cementum annulation analysis by

Matson’s Laboratory (LLC, Milltown Montana).

Besides contemporary samples, 114 samples between the years 1854–1993

obtained from zoological museums in Finland (Oulu, Helsinki and Kuopio) were

utilized (IV). This study material mostly consisted of different kinds of bones:

teeth, some pelvic bones, vertebrae, pieces of skull bone and femurs. Other types

of historical samples included foot pads, nails, pelt samples and dry blood/neural

tissue obtained from inside teeth. The geographic locations for museum samples

are indicated in Figure 5 above.

2.2 DNA extraction, genotyping and sequencing

DNA was extracted with different methods depending on the material type. The

DNeasy® Tissue Kit (Qiagen) or standard phenol-chloroform extraction protocol

was used for tissue samples. The Chelex® method of Walsh et al. (1991) or

DNeasy® Tissue Kit was employed for pelt and hair samples, whereas mouth

swap samples embedded on FTATM cards were extracted according to the protocol

provided by the manufacturer (Whatman). DNA extraction for museum samples

(IV) is presented separately in the following section (2.2.1).

In total, 10–17 different microsatellite loci previously developed for canines

were used throughout this study. These loci were polymorphic and highly variable

in the Finnish and Russian wolf populations (He range ~0.35–0.85), and in

general unlinked and in Hardy-Weinberg equilibrium (I–IV). Significant linkages

detected between loci and deviations from the HW equilibrium in paper III

reflected demographic and genetic changes, e.g. the reduction in the population

census size and the increase in inbreeding.

37

For historical samples (IV), genetic variation in the mtDNA control region

was also investigated. A 431-bp-long target area encompassing the most reported

variable sites of this sequence in wolves was chosen and amplified with PCR

primers developed for this study. Shorter overlapping fragments (~140–180 bp)

were used if no (satisfactory) sequence was obtained with primers covering the

whole target area. PCR products were purified and sequenced using the same

primers, but with a fluorescent dye incorporated. Excess dye was removed

afterwards and the products were run on an ABI 3730 DNA analyser

(PerkinElmer Applied Biosystems).

2.2.1 Handling of museum samples

DNA starts to degrade rapidly after the death of an organism, and over time

shorter and shorter DNA fragments remain and are available for genetic analysis.

The progress of DNA degradation and possibility to extract usable DNA depends

not only on the passing of time, but also, for instance, on the sample type

(Wandeler et al. 2007; Casas-Marce et al. 2010), ambient conditions (e.g.

Martínková & Searle 2006), and preservation methods used on museum

specimens (e.g. Zimmermann et al. 2008).

When the amount of target DNA is low and/or of poor quality, several

precautions must be taken into account in order to avoid contamination and to

ensure the authenticity of the results (see Wandeler et al. 2007 for a review of this

subject). In this study, all pre-PCR phases for museum samples (IV) were

conducted in a special laboratory dedicated to work on old DNA (see IV for

details). DNA extraction from museum bone samples followed a protocol

developed by Rohland & Hofreiter (2007) for old bone samples. Other types of

museum samples were extracted with a DNeasy® Tissue Kit (Qiagen) with some

modifications (see IV for details). In DNA extraction and the following DNA

amplification, several negative controls were run throughout the procedure in

order to detect possible contamination. Additionally, the consistency of obtained

results was confirmed with independent repetitions: For mtDNA sequences the

whole target area of each sequence was covered at least twice. Sequences with

unique or rare haplotypes were amplified up to three more times in both

directions. For microsatellite loci, a heterozygote genotype was not accepted

unless each allele had been observed twice, and a homozygote was not accepted

unless three amplifications were consistently homozygous. If these requirements

were not met after five amplification attempts, half locus or missing data were

38

recorded. Microsatellite amplification and reading of the genotypes was

conducted independently by two persons (Eeva Jansson and Jenni Harmoinen).

2.3 Genetic analysis

As neutral genetic markers were used in this study, the analysis methods and ideas

presented below apply to neutral markers, unless otherwise stated.

2.3.1 Amount of genetic variation

The most commonly used measure to describe genetic diversity is heterozygosity,

which gives the proportion of heterozygous polymorphic loci at the population

level or the proportion of heterozygous individuals for a given locus. Observed

heterozygosity (Ho) is the realized measure, whereas expected heterozygosity (He)

tells the expected proportions based on allele frequencies in the study population

in question (e.g. Allendorf & Luikart 2007, p. 51). The number of alleles (A),

allelic richness (AR), and private allelic richness (PR) are also commonly given

diversity measures, and can be utilized, for example, in bottleneck testing (see

2.3.3) and indirect gene flow indication (see 2.3.5). The amount of allelic

diversity is affected by sampling effort, because rare alleles are more likely to be

included when the population sample is large. Richness measures are adjusted

according to the smallest sample size, so populations with varying numbers of

samples are comparable (Kalinowski 2004). The amount of genetic variation

based on 10–17 polymorphic microsatellite loci, was observed throughout this

study (papers I–IV). In the last paper (IV) where mtDNA sequences were also

utilized, sequence variability was quantified as the number of polymorphic sites

(S), number of haplotypes (H), haplotype diversity (Hd) and nucleotide diversity

(π). Due to differing sample sizes, haplotype and private haplotype richness were

also estimated.

2.3.2 Inbreeding

Inbred individuals and populations have a higher homozygosity and lower

heterozygosity level than expected on the basis of the population allelic

frequencies. This is because the parents of inbred individuals share common

ancestor(s) and are more likely to be autozygous for any given locus (i.e. that

allele is identical by descent; Allendorf & Luikart 2007, pp. 307–308). In the

39

course of this study, we investigated the occurrence of inbreeding on the

population level within the Finnish (I, III, IV) and Russian (II) wolf populations.

The inbreeding coefficients (FIS), i.e. the average departures of genotype

frequencies from Hardy-Weinberg expectations within populations (Wright 1931)

were calculated together with their 95% confidence intervals with an appropriate

resampling scheme. The estimate measures the deficiency or excess of average

heterozygotes in a population sample as:

FIS = 1 − (Ho/He).

2.3.3 Population bottleneck tests

Current genetic methods for detecting bottlenecks from single population samples

are all based on detecting deviations from expectations under mutation-drift

equilibrium, and contrast two different diversity indices, of which one is more

affected by genetic drift than the other (Peery et al. 2012). The ‘heterozygosity-

excess’ test is based on the principle that while heterozygosity is relatively

insensitive to the effects of bottlenecks, allelic variation is easily reduced

(Allendorf & Luikart 2007, p. 127). Rare alleles are more easily lost, and even

more so when the bottleneck is severe, as the probability of a neutral allele being

lost in a bottleneck is:

(1 – p)2N

Where p is the frequency of the allele and N is the size of the bottleneck

(Allendorf 1986). On the contrary, neutral genetic variation measured as

heterozygosity will be lost due to drift in each generation by 1/2N. As rare alleles

are more easily lost in bottlenecks and they contribute only a little to overall

heterozygosity in a population, after a bottleneck the population generally has

excess genetic diversity at selectively neutral loci (i.e. the observed gene diversity

is larger than expected from the number of alleles found in the sample of a

constant-size population; Cornuet & Luikart 1996; Luikart & Cornuet 1998). As a

consequence of rare allele loss, frequency classes based on allele sizes also

become shifted from the normal L-shaped distribution, and significant departures

can be detected with the ‘mode shift test’ (Cornuet & Luikart 1996).

In addition to a deficiency of low frequency allele classes, population

bottlenecks may also generate gaps in the size distribution of microsatellite alleles

(Garza & Williamson 2001), because the number of alleles (K) is expected to

decline faster than the range in allele sizes (r) when there are ≥ 5 alleles in the

40

locus, as by chance more lost alleles are then of intermediate sizes (Peery et al.

2012). This ratio of K/r is the basis of the so-called ‘M-ratio test’ (Garza &

Williamson 2001). The obtained mean ratio across loci from a given dataset is

either (i) compared with ratios derived from putatively stable wild populations, or

(ii) alternatively with simulated distributions under mutation-drift equilibrium

(Peery et al. 2012). Methods based on the principles mentioned above were also

used in this study to detect possible population bottlenecks in the current and

historical Finnish wolf population, and in the neighbouring Russian populations

(I, II, IV).

Immigration and/or mutations can erase the genetic signals of demographic

bottlenecks, which may therefore be detectable only for a short period of time

afterwards (e.g. Busch et al. 2007). In this case, the signal might still be

detectable with non-autosomal genetic markers (mtDNA and cpDNA) as these

sequences are uniparentally inherited and haploid, have only 1/4Ne of that for

autosomal markers such as microsatellites, and are more susceptible to genetic

erosion or the consequences of evolutionary forces (Avise et al. 1984). Population

bottlenecks often create sweep patterns similar to the ones caused by natural

selection. Therefore, neutrality tests originally developed to detect selection, such

as Tajima’s D (Tajima 1989) can be applied even for neutral DNA sequences to

detect past demographic changes. Neutrality tests for variation in the mtDNA

control region were utilized in this study for museum samples (IV).

2.3.4 Population structure

Genetic substructures due to the non-random distribution of genotypes are

peculiar to natural populations. Indeed, many ecological and evolutionary

processes that influence genetic variation are mediated by space (Guillot et al.

2009), and analysis of the spatial genetic structure within continuous populations

in their natural habitat can reveal acting evolutionary processes (Hardy &

Vekemans 1999). Genetic methods may help us to find underlying genetic

structures, but for the correct interpretation of their nature, understanding of the

biology, ecology and demographic history of the study species is crucial.

Two common approaches to study genetic substructures within populations

are to examine whether isolation by distance (IBD) is present, and/or whether the

present genetic variation is divided into distinct, defined clusters (e.g. Guillot et

al. 2009). IBD (Wright 1943; 1946) arises from the limited spatial dispersal of

individuals, and can be seen as a decrease in the genetic similarity (i.e.

41

relatedness) among individuals within populations as the geographical distance

between them increases. As a result of IBD, a population will be divided into

local ‘neighbourhoods’ (Wright 1946), an area from which individuals can be

considered to be drawn at random from a panmictic population (Allendorf &

Luikart 2007, p. 209–210). Neighbourhood size can be simply defined as:

Nb = 4πσ2D,

where σ2 is the mean-squared dispersal distance and D the density of the genes

(Guillot et al. 2009).

Clustering methods look for homogeneous domains by inferring populations

or clusters of individuals that fulfil some genetic criteria (e.g. Hardy-Weinberg

equilibrium, linkage equilibrium or specific allelic frequencies), defining them as

distinct groups (Guillot et al. 2009; François & Durand 2010). Unlike methods

such as F-statistics (see 2.3.5), which rely on predefined subpopulations, these

Bayesian methods examine whether and how many of such substructures (i.e.

clusters) could be found from data with or without individual spatial information

included (François & Durand 2010).

In this study, the patterns of IBD (I, III) and genetic clusters (I–IV) within the

Finnish and Russian wolf populations were investigated. STRUCTURE

(Pritchard et al. 2000), the most popular clustering approach, was utilized in all

studies (see attached papers for details). In paper I, another Bayesian approach

(BAPS; Corander et al. 2003) was also used. The pattern of isolation-by-distance

and the size of neighborhood were estimated only within the contemporary

Finnish wolf population, however (I, III).

2.3.5 Population connectivity and gene flow

Genetic methods offer a practical means to assess population connectivity,

because it is often difficult to measure dispersal directly at large spatial scales

(Lowe & Allendorf 2010). Most commonly used methods to describe the

distribution of genetic diversity between the levels of hierarchy (e.g. individuals,

subpopulations and total population) are based on Wright’s F-statistics (Wright

1931; see also Excoffier et al. 1992). If subpopulations are genetically diverged,

the proportion of genetic diversity due to allele frequency differences among

populations, FST will be > 0 (Holsinger & Weir 2009). This divergence between

subpopulations (or any predefined levels of structures) refers to a limited amount

of dispersal and gene flow among them, and the effect of genetic drift changing

42

the allele frequencies. In this study, genetic differentiation was examined between

the Finnish and Russian wolf population (II) in order to determine whether the

amount of gene flow is limited, but also between temporal samples among the

Finnish population (III, IV) to see, if the local Ne is small enough for genetic drift

to cause genetic divergence within the studied time frame.

Genetic differences were further investigated by factorial correspondence

analysis (FCA; III, IV), which calculates the maximum genetic variation among

individuals, generates axes that describe this variation and then plots individuals

along these axes according to their genotype. Graphs from such plots offer a

convenient way to illustrate genetic differences between populations.

Additionally, indirect and direct genetic methods (reviewed in Lowe &

Allendorf 2010) were used to estimate gene flow between Russian and Finnish

wolf populations (I, II, III). The average number of migrants, Nm between Russian

populations and Finnish population was estimated in papers II and III with the

private allele method of Slatkin (1985; Barton & Slatkin 1986), which is based on

the assumption that in genetically subdivided populations the logarithm of

average frequency of private alleles is approximately linearly related to Nm. The

proportion of immigrants, m, was quantified in papers II and III with a Bayesian

method (Wilson & Rannala 2003) that calculates inbreeding coefficients for each

population, the joint probabilities of which are then used to estimate recent

migration rates (paper II), or in conjunction with temporal Ne estimation (paper

III) using the likelihood-based method by Wang & Whitlock (2003).

Assuming that subpopulations are genetically diverged, likely immigrants

may be identified directly on the basis of their multilocus genotypes. This can be

done, for example, by Bayesian assignment analysis (Rannala & Mountain 1997),

which was utilized in papers I and II. In paper I, in which no likely source

population samples from Russian Karelia were available, assignment of

immigrants was based on a self-classification approach (i.e. individuals with

genotypes not fitting the study population are possible immigrants). In paper II,

with Russian Karelian population samples included, the migration rate between

these two populations was estimated. Each individual’s multilocus genotype was

compared to the marginal probabilities from randomly generated distributions,

which determines the statistical probabilities belonging to the tested populations.

43

2.3.6 Effective population size

Methods to estimate Ne can be divided on the basis of the number of population

samples required for (i) point and (ii) temporal estimators (single population

sample vs. two or more samples of the same population at different sampling

times, respectively). Temporal methods are based on the premise that genetic drift

in neutral allelic frequencies is inversely proportional to the effective size (e.g.

Palstra & Ruzzante et al. 2008), i.e. they measure the harmonic mean of the

variance effective size (NeV) over the interval between samples (Waples 2005). On

the contrary, single-sample methods estimate in general the inbreeding effective

size (NeI) of the parental generation or recent past (Waples & Do 2008; Barker

2011) from different genetic factors related to Ne, including linkage

disequilibrium (see Luikart et al. 2010 for a review).

Besides applying to different time frames, and never for the last sampled

population itself (Waples 2005), the two type of effective sizes presented above

are expected to be roughly the same only when the population has reached an

equilibrium state, and if underlying demographic and ecological dynamics have

remained constant for a long time (Palstra & Ruzzante 2008). As this is hardly

ever the case in natural populations, it is clear that interpreting the estimates from

different approaches can be challenging but necessary and valuable for the

interpretation of the results. Moreover, several assumptions that accompany

genetic Ne estimation methods are often violated and may create biased estimates

(reviewed in Luikart et al. 2010). In particular, violating the assumptions of

isolation, a stable population size and discrete generations may be difficult or

impossible to avoid in many sampling schemes (including this study), and are

thus likely to affect the estimation (e.g. Palstra & Ruzzante 2008; Luikart et al.

2010; Barker 2011). Possible biases that might have affected the obtained

estimates in this study are discussed in the attached original papers (I–IV).

In the course of this study, effective population sizes (and its changes) in the

contemporary Finnish and Russian wolf populations were estimated using several

methods. In paper I, samples collected in 1996–2004 were used to estimate the

contemporary and historical Ne (late 19th or early 20th century), in paper II single-

sample Ne estimates were obtained for the Karelian and Archangel wolf

populations, and in paper III contemporary Ne and its variation within the Finnish

wolf population during large demographic changes (1995–2010) was monitored,

whereas in paper IV, Ne and its temporal variation was estimated from museum

samples collected during the last ~150 years. In total, nine methods to estimate Ne

44

and its changes were used in this study (two single-sample methods, 6 temporal

methods and one simulation method for the detection of demographic changes).

Due to their large number, these methods are presented in the original papers (I–

IV) only.

45

3 Results and discussion

Papers I and III somewhat overlap, because in both studies the contemporary

genetic diversity and population structure of the Finnish wolf population in 1996–

2004 were considered. All wolves analysed in paper I were also included in paper

III. However, the time frame in paper III was until 2009, and thus it included the

demographic crash in the wolf population after 2006. Additionally, wolves in

paper III were grouped into temporal cohorts (corresponding approximately to the

generation interval of wolves) based on their estimated or known year of birth

instead of the sampling date (as was carried out in paper I). Because generational

overlap may bias the estimates of genetic characteristics, this division is likely to

give more accurate results on the temporal variation in them (e.g. Luikart et al.

2010). Therefore, the results of paper III are prioritized when the two studies

overlap.

3.1 Amount of genetic variation

Despite the documented historical demographic bottlenecks in the Finnish and

neighbouring north-western Russian wolf populations (e.g. Pulliainen 1965; 1980,

Flagstad et al. 2003; Danilov 2005), the amount of genetic diversity measured in

the means of heterozygosity and allelic diversity in the different studies was

relatively high (Table 1). A varying number of loci (10–17) were used in different

periods and geographic areas. Thus, the results presented in the separate papers

are not perfectly cross-comparable or comparable with other studies due to the

use of different loci. However, in general the levels of heterozygosity were

relatively stable (Table 1), and well in line with earlier studies, in which Finnish

wolf samples have been used for reference purposes: Flagstad et al. (2003)

reported Ho of 0.69 (± 0.12) and He of 0.72 (± 0.09) for the ‘contemporary

Finnish wolves’ (N = 22) and in Lucchini et al. (2004) with 13 Finnish wolves

analysed Ho was 0.69 (± 0.13) and He was 0.73 (± 0.07). The heterozygosity

levels among modern Finnish wolves (I, III) resemble those reported in much

larger, non-endangered wolf populations (cf. vonHoldt et al. 2010; Sastre et al.

2011).

Some interesting temporal changes of diversity were observed, however: The

expected heterozygosity in the modern Finnish wolf population (III) varied from

0.677 to 0.709, being highest among individuals born in 1995–1997 and lowest at

the end of study period (2007–2009). Observed heterozygosities followed a

46

similar trend, and were 0.749 in 1995–1997 and 0.615 in 2007–2009 (Table 1).

The decrease in Ho was statistically significant (P = 0.021), and was probably due

to the population crash after 2006, increased inbreeding (see chapter 3.2) and the

low amount of efficient gene flow (3.5) after the population recovery. At the

beginning of the study period, the Finnish wolf population census size was small

and there were only a few breeding individuals (Fig. 3), but the genetic diversity

was still at its highest. Thus, it is likely that the high diversity reflected the

genetic diversity of the Russian Karelian source population on which the Finnish

wolf population was demographically dependent at that time (e.g. Kojola et al.

2006). Indeed, gene diversity estimates (He) for the Karelian samples collected in

1995–2000 (paper II) and Finnish samples born in 1997 or earlier (‘Finnish 1995–

1997’ in paper III) were identical (0.709).

Table 1. Combined results from papers II–IV showing the mean diversity indices and

inbreeding coefficients (FIS; see 3.2) within the contemporary Finnish and Russian

wolf populations and in the temporal Finnish museum groups. The number of

microsatellite loci used and samples analysed (N) are shown. The neighbourhood size

(Nb) estimates presented are covered in chapter 3.4.3.

Population sample (paper) Number

of loci

N He Ho FIS A AR Nb

Arkhangelsk 1995–2000 (II) 10 14 0.636 0.634 0.051 4.7 4.2 NA

Karelian 1995–2000 (II) 10 29 0.709 0.656 0.094* 5.7 4.6 NA

Karelian 1995–2010 (III) 17 37 0.733 0.691 0.074 6.3 NA NA

Finnish 1995–1997 (III) 17 43 0.709 0.749 −0.044* 6.5 5.9 131.4

Finnish 1998–2000 (III) 17 51 0.684 0.673 0.028 5.9 5.4 49.8

Finnish 2001–2003 (III) 17 83 0.695 0.686 0.030 6.5 5.8 71.6

Finnish 2004–2006 (III) 17 87 0.687 0.693 −0.002 6.2 5.6 76.3

Finnish 2007–2009 (III) 17 33 0.677 0.615 0.108* 6.0 5.7 31.5

Finnish before 1920 (IV) 15 12 0.669 0.624 0.113 5.0 3.7 NA

Finnish 1920–1959 (IV) 15 6 0.721 0.741 0.079 4.7 4.2 NA

Finnish 1960–1979 (IV) 15 22 0.686 0.729 −0.038 5.5 3.6 NA

Finnish 1980–1993 (IV) 15 18 0.676 0.622 0.112 5.5 3.7 NA

Finnish 1995–2009 (IV) 15 30 0.697 0.712 −0.003 5.9 3.7 NA

He, expected heterozygosity, Ho, observed heterozygosity, A number of alleles, AR, allelic richness.

* Statistically significant (P < 0.05) deviations in FIS values. NA, not analysed

Genetic diversity measured as the number of alleles and allelic richness did not

reveal large temporal changes in any of the studies (Table 1). In the study

concerning historical genetic variation (IV), an interesting pattern was

47

nevertheless observed: Almost 20% of the alleles (21/107) present in museum

samples (collected in 1854–1993, N = 58) were not found in the pooled modern

population sample (1995–2009, N = 298), whereas only four alleles (3.7%) were

unique to the concurrent population. The high frequency of these ‘ghost alleles’ in

museum samples suggests that the historical wolf population was in fact more

diverse, and that allelic variation has been lost in demographic bottlenecks (see

e.g. Nyström et al. 2006; Ugelvik et al. 2011 for similar observations from

Scandinavian arctic foxes and Danish large blue butterflies, respectively). In order

to see substantial changes in the level of heterozygosity, rather special

circumstances may be required, such as a long-term small effective size without

immigration (see e.g. studies with isolated wolf populations from Scandinavia;

Vilà et al. 2003 and from Italy; Lucchini et al. 2004), because even in the tightest

bottleneck with one breeding pair surviving, ~75% of the heterozygosity is

expected to remain in the following generation (e.g. Allendorf & Luikart 2007, p.

127). On the other hand, rare alleles are easily lost in a population bottleneck (see

2.3.3), and they are therefore likely to be more sensitive indicators of past

demographic changes, even with some continued gene flow.

Mitochondrial sequences are more prone to genetic erosion in conjunction

with demographic fluctuations than autosomal markers due to their lower

effective size and mode of inheritance (see 2.3.3). Only three haplotypes have

been found in the modern Finnish wolf population (Kiiskilä 2006), and these are

the same ones that were among the most recent historical wolves collected in

1980–1993 (Table 2 below). In total, eight haplotypes were observed in the

historical wolf population. Five of them were rare (see IV for details) and have

most likely been lost in the modern population. Because the lost alleles were rare,

the frequency of private alleles was considerably higher in the samples collected

before 1960, and the number of polymorphic sites has also reduced notably (Table

2). Similarly to nuclear genetic markers, the change in diversity (Hd) was smaller

(and not significant).

48

Table 2. Mitochondrial diversity indices of Finnish wolves in different time periods

(Table included in paper IV).

Temporal sample N H HR PR Hd π S

Before 1920 18 5 2.80 1.13 0.556 ± 0.130 0.0070 15

1920–1959 6 3 3.00 2.00 0.600 ± 0.215 0.0132 14

1960–1979 29 4 2.35 0.25 0.554 ± 0.064 0.0108 11

1980–1993 28 3 2.17 0.18 0.421 ± 0.103 0.0086 10

N sample size, H number of haplotypes, HR haplotype richness, PR private haplotype richness,

Hd haplotype diversity, π nucleotide diversity, S number of polymorphic sites

3.2 Inbreeding

Differences between the expected and observed heterozygosities resulted in a

significant positive inbreeding coefficient among Finnish wolves in the period

2007–2009 and among the Russian Karelian wolves collected in 1995–2000

(FIS = 0.108 and 0.094, respectively; Table 1). By contrast, the inbreeding

coefficient was significantly negative in the Finnish population in the course of

population recovery in 1995–1997 (FIS = −0.044), suggesting active avoidance of

inbreeding.

A positive inbreeding signal obtained indirectly from F-statistics as

heterozygosity deficiency could also be due to excess sampling of (closely)

related individuals (e.g. Jankovic et al. 2010) or a cryptic population substructure

(i.e. Wahlund’s effect; e.g. Allendorf & Luikart 2007, p. 202). However, the

average relatedness (r) between sample pairs in the contemporary Finnish wolf

populations was confirmed to be generally low (range −0.029 to 0.026; see III for

details), and no significant spatial substructures were found within study

populations [no subdivision was detected in the Russian Karelian and Archangel

wolf populations, (II) and genetic clusters detected in the Finnish population

probably reflected family lines with no clear spatial patterns (I, III)]. Therefore, it

is likely that the observed significant inbreeding coefficients in this study reflect

true changes in inbreeding, e.g. due to population size decline, a low amount of

gene flow and disturbances in pack structures (III).

In summary, inbreeding has recently increased in the Finnish wolf population,

and the inbreeding coefficient for the individuals born in 2007–2009 was

relatively high (FIS = 0.108; Table 1) and similar to those reported in isolated wolf

populations in Europe (e.g. Italy FIS = 0.127; Verardi et al. 2006; Iberia

FIS = 0.177; Sastre et al. 2011). Due to well-known detriment of inbreeding

49

depression to fitness and population viability (see Chapter 1.5), the recent trend is

worrying, even though no signs of inbreeding depression have yet been detected

in Finnish wolves. On the other hand, no such signs have been looked for, either.

3.3 Population bottlenecks

Bottleneck tests of the modern (I) and historical (IV) Finnish wolf populations as

well as those of the Russian populations (II) gave somewhat inconclusive results

(Table 3).

Table 3. Compiled results of the population bottleneck tests in this study. Numbers

inside parentheses after each group refer to the paper in which a given sample was

analysed. Note that the significance of M-ratio tests in paper IV was based on

simulated distributions of the ratio, whereas in papers I and II bottlenecks were

compared to a critical ratio of 0.68. Therefore, these results are not fully comparable.

Population sample Heterozygosity

excess test

Mode-shift

test

M-ratio test Tajima’s D Fu’s FS

Finnish 1996–1998 (I) -- -- -- NA NA

Finnish 1999–2001 (I) -- -- -- NA NA

Finnish 2002–2004 (I) + -- -- NA NA

Finnish before 1920 (IV) -- -- + -- --

Finnish 1920–1959 (IV) --1 +1 NA --1 --1

Finnish 1960–1979 (IV) -- -- -- + +

Finnish 1980–1993 (IV) -- -- ? -- +

Finnish 1995–2009 (IV) -- -- ? NA NA

Karelian 1995–2000 (II) -- -- -- NA NA

Arkhangelsk 1995–2000 (II) -- + -- NA NA

+ Indicates a preceding population bottleneck, -- indicates no support for a preceding bottleneck, NA not

analyzed, ? preceding bottleneck possible depending on the evolutionary model used, 1note the small

sample size

The heterozygosity excess method suggested a pre-existing bottleneck only

among the Finnish wolves in 2002–2004. Moreover, mode shift inspection

revealed an allelic frequency distribution typical of bottlenecked populations in

the historical small sample (N = 6) collected in 1920–1959 and in the modern

Arkhangelsk wolf population, whereas M-ratio simulations supported Ne decline

in the oldest museum samples (before 1920) under most evolutionary scenarios

(see IV for details). On the other hand, demographic tests for mitochondrial

sequences, which were conducted for museum data only, showed significant

50

signals of diversity loss in the wolf population in 1960–1993. None of the

conducted tests suggested bottlenecks in the Karelian wolf population, nor in the

Finnish wolf population collected during and directly after the population

recovery (1995–2001).

Different moment-based tests often fail to detect bottlenecks, even in cases

where large reductions of population size are known to occur (e.g. Busch et al.

2007; Peery et al. 2012). There are many possible explanations for this

contradiction: i) a bottleneck may not be severe enough to leave a genetic

bottleneck signal, ii) a bottleneck may have happened too long a time ago, iii)

immigration may have erased the genetic bottleneck signal, iv) sample size may

have been too small, and/or v) the assumed evolutionary parameters may have

been incorrect (Peery et al. 2012). In the Finnish and Russian wolf populations,

gene flow (see chapter 3.5) could have been the most important factor

overpowering the bottleneck signal(s). Evolutionary parameters for the used

microsatellite loci are also largely unknown and sample sizes rather modest,

which could also have affected the results.

Moreover, methods differ in their sensitivity depending on the timing and

severity of the bottleneck (Williamson-Natesan 2005) and on the pre-bottleneck

level of genetic diversity (Peery et al. 2012). For instance, the M-ratio test is

likely to detect a population bottleneck better than the heterozygosity excess test

when the bottleneck has been severe and long-lasting (Williamson-Natesan 2005),

and has more statistical power when the pre-bottleneck genetic diversity was high

(Peery et al. 2012). However, when the underlying evolutionary model of the

used microsatellites and/or the historical effective population size are not known,

simulation studies to determine deviant M-ratios might be needed (Peery et al.

2012), and the often-used limit ratio of ~0.7 derived from putatively stable wild

populations (thus ratio lower than that would indicate a likely bottleneck; Garza

& Williamson 2001) might not always apply (e.g. Busch et al. 2007; Ugelvik et

al. 2011). Despite a relatively high M-ratio of 0.794 among the Finnish wolves

sampled before 1920, a bottleneck was supported in this group under the most

realistic evolutionary scenarios (see IV for details). In other temporal museum

groups, with an unlikely bottleneck, the M-ratios were clearly higher, 0.856–

0.910, as was the case in the concurrent Russian populations (0.850 in Karelia and

0.900 in Archangel; II). Therefore, it is possible (but remains to be confirmed)

that the M-ratios of 0.71–0.73 measured in the Finnish wolf population in 1996–

2004 (I) would indeed also indicate genetic bottlenecks.

51

3.4 Population structure

Regional and continental patterns of genetic subdivision are generally found even

in the grey wolf, a species with extremely high dispersal abilities (e.g. vonHoldt

et al. 2011, Scandura et al. 2011; Stronen et al. 2012). This study demonstrated

that the present-day Finnish wolf population did not form a totally panmictic unit

within Finland (I, III, IV), but consisted of varying number of clusters, was

characterized by IBD (I, III) and was genetically distinct from the neighbouring

Russian population (chapter 3.5; II, III).

3.4.1 Cryptic population structure – clustering analyses

The number of detected clusters (K) in the modern Finnish wolf population varied

temporally (K = 3–8) following population size trends (see III for details). The

most plausible biological explanation for these clusters and their varying number

was that they represent ‘family lines’ (I and III), whose number was smaller than

the number of wolf packs (Fig. 3; number of breeding pairs is equivalent to the

number of packs) due to the limited number of breeders and shared ancestry of

many individuals and wolf packs. Groups of closely related individuals

generating clusters in Bayesian population structure analyses are often seen as a

problematic artefact (e.g. Rodríguez-Ramilo & Wang 2012), but can in fact be

used to reveal underlying population core structures and changes in them in small

populations characterized by close kin groups. The detected clusters in this study

did not clearly overlap spatially with known wolf pack territories, and in some

cases the clusters had a very wide geographic distribution, suggesting frequent

migration between wolf packs and non-existing spatial genetic hierarchy (III).

The number of family lines might have been overestimated however, because

Bayesian clustering methods tend to be upward biased (i.e. include false

positives) when substructures are found within populations characterized by IBD

(e.g. Frantz et al. 2009; Meirmans 2012; see 3.4.2).

Similar within-population substructures have been detected, for instance,

among wolf populations of the Northern Rocky Mountains in North America

(vonHoldt et al. 2010) and in the Italian Apennines (Scandura et al. 2011), and

they might be a more general phenomenon reflecting the limited number of

founders in (re)established populations (this study), the high stability of

established packs, and/or short mean dispersal distances (vonHoldt et al. 2010;

Scandura et al. 2011; Randi 2011). Anthropogenic perturbations are likely to

52

reinforce this genetic divergence within and also between populations, if

dispersing individuals are eliminated and hence connectivity is reduced (vonHoldt

et al. 2010; Stronen et al. 2012).

The cluster analysis of the temporal groups (IV) also revealed internal

population structures within the historical Finnish wolf population and changes in

their proportions during the study period. Temporal clustering is possible when

each of the temporal groups represents long time periods, and genetic drift causes

genetic temporal differentiation (e.g. Flagstad et al. 2003). All temporal museum

groups were significantly differentiated (FST = 0.033–0.096; IV), and the pairwise

genetic distances between groups were positively correlated with their median

sampling year (Mantel's test: r = 0.838, P = 0.048). However, genetic drift is not

probably the sole cause for the detected subgroups, because the assignment

probabilities of the wolves in different temporal groups to the clusters did not

decrease linearly with the time difference. The biological significance of the three

groups discovered is hardly related to explicit family structures either (due to the

long study period), and remained somewhat unclear. Interestingly, however, one

of these clusters was peculiar to the oldest samples (before 1920), and all

individuals within this group were from Northern Finland (see Fig. 9 in paper IV).

A distinction between the genetic composition of the northern and southern

wolves was also discovered in the study of Scandinavian historical wolves by

Flagstad et al. (2003). Thus, it is possible, that some wolf type typical of Lapland

was lost when wolves from these areas were effectively eliminated because of

their threat to reindeer herding.

3.4.2 Limited dispersal – IBD

Spatial autocorrelation analyses revealed a clear, significant pattern of isolation

by distance within the concurrent Finnish wolf population (I, III), indicating that

individuals close to each other geographically were on average more related than

those further apart from each other. In general, deviations from population mean

kinship estimates were largest in the closest (< ~20–100 km) and average distance

classes (~200–300 km), but smaller in the long distance classes (> 400–600 km).

These long distance classes probably represent those few wolves that dispersed

furthest from their natal packs and colonized new, uninhabited areas (I), and are

therefore important for the demography and spatial dynamics of the population

(e.g. vonHoldt et al. 2010; Randi 2011) but irrelevant with respect to the IBD

concept (cf. Vekemans & Hardy 2004). The slope of the distance regression

53

varied temporally (III; Fig. 6 below): the negative slope of the regression was

smaller (b = −0.007, P = 0.006) in the temporal groups at the beginning of the

study period (1995–1997) than in the later periods and specifically in the last one

(2007–2009; b = −0.028, P = 0). Consequently, neighbourhood size (Nb)

estimated from the slope and the average kinship between adjacent individuals

(see I) was much larger (131.4) in the temporal group representing the population

recovery phase (1995–1997) than in the group at the end of the study period

(2007–2009; Nb = 31.5; see Table 1 for estimates from other time periods). The

average dispersal distance estimated from samples collected in 1996–2004 (I) was

97.2 km, and well in line with a field study by Kojola et al. (2006), in which the

dispersal of wolves equipped with radio or GPS transmitters (N = 60) was

followed, and the median dispersal distance was 98.5 km (range 35–445 km).

Fig. 6. Kinship coefficient vs. distance (log scale) between individuals in the first and

last temporal group included in this study. Asterisks denote a significant deviation

from the population mean (* P < 0.05, ** P < 0.01, *** P < 0.001). Figures were included

as electronic Appendices in paper III.

The genetic population structures of many animal species are often greatly

affected by different social bonds (e.g. Pope et al. 2006; Hoelzel et al. 2007;

Gobush et al. 2009). The model of IBD in highly mobile wolves also stem from

sociality: established packs are usually very large in size (see 1.1.2) and spatially

54

stable (alpha wolves do not leave their home territory; Kojola et al. 2006, see also

Scandura et al. 2011). Because the mean dispersal distances of young wolves

leaving their natal packs are also relatively short (this study; Kojola et al. 2006),

the pattern of IBD is not surprising. Short average dispersal distances in relation

to high mobility are probably a consequence of several factors. The abundance of

prey and available territories in areas near the natal territory favour short dispersal

distances (e.g. Taylor & Norris 2007), whereas population expansion (Randi

2011), a high local density and inbreeding avoidance may trigger long-range

dispersal events (e.g. Scandura et al. 2011). The clear steepening of the IBD

during the heavy decline in the Finnish wolf population (2007–2009; Figs. 3 and

6) was concurrent with the increase in inbreeding (Table 1). One mechanism that

could explain this phenomenon is the sudden loss of alpha individual(s) close to a

breeding season (Brainerd et al. 2008). Consequently, this would increase the

internal turnover rate of wolf packs and the likelihood of close inbreeding, and

lead to higher genetic similarity among individuals within short distances.

3.5 Population connectivity and gene flow

The maintenance of dispersal between subpopulations is critical for endangered

large carnivores with social pack structures and possessing large home territories

(Marsden et al. 2012), because very large population sizes may be needed to

ensure the viability in the long term (e.g. Reed et al. 2003; Traill et al. 2007,

2010). Accordingly, for the Finnish wolf population, connection with the Karelian

wolf population is of key relevance.

3.5.1 Genetic differentiation between Finnish and Russian wolves

Several analyses demonstrated that the Finnish and north-western Russian wolf

populations do not nowadays form just one panmictic population, but are

genetically differentiated. In the analyses included in paper II, the genetic

distances were relatively large: FST was 0.151 between the Karelian and Finnish

wolf populations and 0.176 between the Arkhangelsk and Finnish populations.

This amount of divergence is very close to that (0.177) reported by Seddon et al.

(2006) in their study on Finnish and Scandinavian wolf populations. The latter is

located far away (> 600 km) and has been effectively isolated since the

population re-establishment in the 1980s due to the geographical distance and the

reindeer herding areas, being almost impenetrable for wolves (e.g. Seddon et al.

55

2006; Hansen et al. 2011). Based on AMOVA analysis, a considerable proportion

(15.2%) of the total genetic variation was between the wolf populations of Russia

and Finland. These three populations were also clearly separated into their own

clusters in STRUCTURE analysis (see II for more details). Later inspection of the

genetic differentiation between the Finnish and Karelian population (III) did not

suggest any major changes: FST was somewhat lower (0.086) between the

contemporary Finnish (2007–2009) and Karelian (2009/2010) wolf populations,

but the differentiation was still highly significant and the amount of gene flow

between the populations has remained low (see below).

3.5.2 Amount of gene flow and detection of immigrants

The fairly high genetic differentiation detected between the Russian and Finnish

wolf populations indicates a low number of immigrants between them. According

to Wright’s island model (Wright 1951), an FST value of 0.152 is equivalent to

~1.4 effective migrants per generation between populations (II), whereas the later

measured FST of 0.086 (III) corresponds to ~2.7 migrants per generation. Another

indirect method based on private allele frequencies showed a reverse trend and

earlier suggested 3.0 migrants per generation (Nm) between the Russian Karelian

and Finnish wolf populations (II), although when the estimation was repeated at a

later stage and with all available wolves in both populations (III), Nm was only

1.09. Nevertheless, both of these methods suggested a low migration rate in the

range of ~1–3 immigrants per generation between the populations. Indirect

methods may not be very reliable, however, because they assume equilibrium

between drift and migration and their assumptions are overly simple for most wild

populations (e.g. Whitlock & McCauley 1999). Thus estimation of the proportion

of immigrants (m) with other approaches might be more robust (Waples &

Gaggiotti 2006; Lowe & Allendorf 2010).

The estimates obtained with Bayesian methods also suggested relatively low

migration rates between the populations: the first estimate using the Wilson &

Rannala (2003) approach suggested an m value of 0.069 (± 0.032 SD) from the

Karelian into the Finnish wolf population (II), whereas m estimated jointly with

Ne (Wand & Whitlock 2003) for the different periods in paper III varied from

0.016 (2001–2006) to 0.090 (1998–2003; see III for details). Based on the

assignment of individual genotypes, the self-classification approach (I) suggested

that only four wolves were probable first-generation immigrants (3%), whereas in

the later study (II) with the Karelian population used as a reference, the overall

56

migration rate between the population was 2.7% (m = 0.03), but only one

individual was a probable immigrant into Finland. No individual-based

assignments were conducted in the latest temporal period, but FC analysis

identified one likely immigrant wolf from Karelia into Finland in 2008 (see III for

details).

The findings of relatively high divergence and a low amount of migration

between the Finnish and Russian wolf populations has significant implications

concerning the management of the Finnish wolf population: Contrary to previous

conceptions (Pulliainen 1965, 1980; Boitani 2003; Linnell et al. 2008) and

common assumptions, the connection between the wolf populations seems to be

weak on the basis of this study.

There are several possible explanations for the low amount of wolf

immigration to Finland (see II for details), of which the dependence on local

densities (e.g. Taylor & Norris 2007) during the dispersal phase might be the most

important. If local population densities are low and territories are available near

the natal pack territory, there is no ‘need’ for long-range dispersal (see 3.4.2).

Wolves are not protected in Russia, and in the Karelian population, for example,

the number of wolves has fluctuated greatly, a recent estimate being only around

300–350 individuals (see II for more details and references). Thus, the local

density in Karelia is likely to be rather low, and might reduce the migration

tendency within Karelian wolves. Even a very low amount of immigration could,

however, preserve the adaptive variation (Nm > 0.1) and enable the avoidance of

inbreeding (Nm > 1), but not prevent the differentiation of allele frequencies

between subpopulations due to genetic drift. Importantly, immigrants must also

breed in order to have any genetic significance in the recipient population (e.g.

Lowe & Allendorf 2010). Probable immigrants detected in this study were all

killed or found dead near the Finnish–Russian border (see I and III for details). It

is possible that some immigrants were not sampled in this study, and that these

individuals could have even gained a reproductive alpha status. However, because

no significant positive changes in the Finnish gene pool were noticeable and the

inbreeding coefficient actually increased among wolves born by the end of the

study period (2007–2009; Table 1), it is likely that the amount of effective gene

flow is currently lower than the 1–2 wolves per generation required to ensure the

long-term survival of the population (Anonymous 2005).

57

3.6 Effective population size

Single-sample-based estimates of the effective population size in the current

Finnish (I, III) and Karelian (II) wolf populations were relatively small (20.4–

76.4; Table 4) throughout this study. These approaches are widely used (e.g.

Luikart et al. 2010) and they have been proven to be quite accurate when the

actual Ne is small (e.g. Waples & Do 2008; Barker 2011; Hoehn et al. 2012). LD-

based estimates of Ne are prone to biases if other factors than a small effective

size, such as the substructure, are behind LD (Luikart et al. 2010). However, they

appear to be rather insensitive to the effect of gene flow even with an m of ~0.10–

0.25, and thus may be used in detecting early indications of population

fragmentation and decline (England et al. 2010; Luikart et al. 2010) and they may

provide reasonably reliable estimates of local Ne (Waples 2010). The ONeSAMP

method (Tallmon et al. 2008) uses seven summary statistics besides LD in its Ne

estimation and might therefore provide more precise estimates (e.g. Luikart et al.

2010). Estimates derived from ONeSAMP seem to be sensitive to sample size,

however (Sotelo et al. 2008; Haag et al. 2010). This effect was also confirmed in

this study as smaller random data sets from the original sample provided

significantly smaller estimates in two out of five cases (see III for details).

Thereby, estimates based on small population samples often characteristic of

studies of endangered populations, such as those in paper (IV) (Table 4), might

also be downward biased.

Table 4. Compiled results for Ne estimates based on single-sample approaches in this

study. Estimates from an LD-based method and from a Bayesian method with their

95% confidence intervals or limits are shown. Roman numerals after each group refer

to the original articles.

Population sample N LD-Ne 1 95% CI ONeSAMP2 95% CI

Karelian 1995–2000 (II) 29 46.7 38.2–115.8 39.9 24.8–80.0

Finnish 1995–1997 (III) 43 67.2 54.5–85.2 35.8 28.8–52.2

Finnish 1998–2000 (III) 51 27.1 23.3–31.5 29.3 24.0–45.0

Finnish 2001–2003 (III) 83 30.3 27.0–34.1 31.7 25.6–47.2

Finnish 2004–2006 (III) 87 37.4 33.4–42.0 51.5 44.4–78.3

Finnish 2007–2009 (III) 33 20.4 17.6–23.8 30.4 25.4–45.4

Finnish before 1920 (IV) 12 20.5 13.5–35.2 13.2 10.8–19.6

Finnish 1960–1979 (IV) 22 76.4 48.1–159.0 24.3 20.9–32.7

Finnish 1980–1993 (IV) 18 45.2 30.4–78.4 23.1 18.5–36.3 1Waples & Do 2008, 2Tallmon et al. 2008

58

More thorough investigation of the more robust LD-based estimates revealed

interesting patterns: The effective population size was relatively high during the

population recovery phase (1995–1997; Ne = 67.2), and the wolf population was

characterized by concurrently high genetic diversity, inbreeding avoidance and a

large neighbourhood size (Table 1), probably reflecting the genetic composition

of the Russian source population at that time (III). After this period, there was

probably a population bottleneck, when only a proportion of the individuals

reproduced (Fig. 3), and thus estimates for later time periods (1998–2009) are

significantly smaller (Table 4).

The estimate of Ne for the Karelian population in 1995–2000 was somewhat

smaller and less precise than that for the Finnish wolves born before 1998. This

might be, for example, due to the longer sampling interval and because an

unaccounted age structure tends to downwardly bias the estimates (Palstra &

Ruzzante 2008). Regardless, the estimate for the Karelian population is

surprisingly low. In a recent study by Sastre et al. (2011), 47 Russian wolves

south-east from our study area in the regions of Tver, Vologda, Smolensk and

Kaluga Oblasts (covering an area > 300 000 km2) were genotyped and the

population Ne estimated. Their analysis suggested an LD-Ne of 138.0 (CI 75.9–

490.4) for the wolf population of this continuous area, which together with the

detected high inbreeding coefficient (FIS = 0.147, CI 0.07–0.20) might suggest

that the Russian wolf population is nowadays somewhat fragmented, contrary to

what is commonly assumed. The Ne estimates from the Finnish and Karelian wolf

populations obtained in this study support this hypothesis.

Temporal approach estimates revealed a considerable temporal fluctuation of

Ne within the Finnish wolf population (Table 5 below; III). When compared to the

single-sample estimates (Table 4) and to the estimated number of breeding

individuals (Fig. 3), these estimates were in many cases extremely high, and thus

biologically implausible. It is likely that a longer time between temporal

population samples could have given more accurate and precise results, because

the drift signal against background noise would have been stronger (e.g. Palstra &

Ruzzante 2008; Waples 2010; Luikart et al. 2010). For the monitoring of

endangered populations, the recommended longer sampling interval of many

generations is impractical (e.g. Waples 2010), and estimates from a temporal

approach are thus of little value. Estimates based on this approach derived from

historical population samples can be of considerable value in conservation,

however, because they can provide a requisite base estimate of Ne in a previous,

undisturbed population.

59

Table 5. Combined Ne estimates from three different methods based on the temporal

approach together with 95% confidence intervals or limits for each estimate. Roman

numerals after each time period refer to the original articles.

Time period Moment based

method1

Coalescent Bayesian

method2

Pseudo-ML method3

1995–1997 to 1998–2000 (III) 15.0 (8.0–118.0) 32.1 (18.0–74.7) 46.0 (28.9–104.0)a

1998–2000 to 2001–2003 (III) 17.0 (12.0–30.0) 38.0 (23.7–71.1) 35.5 (25.1–56.3)a

2001–2003 to 2004–2006 (III) 97.0 (47.0–∞) 159.1 (59.3–∞) 200.2 (70.6–∞)a

2004–2006 to 2007–2009 (III) 39.0 (23.0–127.0) 84.2 (34.5–∞) 79.7 (39.8–∞)a

1854–1914 to 1980–1993 (IV) 86.0 (63.0–139.0) 176.4 (119.8–283.0) 153.2 (102.3–249.1)b

1 TempoFS; Jorde & Ryman 2007, 2 CoNe; Anderson 2005, 3 MNE; Wang 2001, Wang & Whitlock 2003, aImmigration source from the Russian Karelia included, bpopulation assumed to be closed.

Effective size estimates for the historical Finnish wolf population based on

temporal approaches and using an interval of ~25 generations were relatively

large, varying from 86.0 to 176.4 (Table 5; IV). Instead of the local Finnish wolf

population, long-term estimates are likely to reflect the genetic behaviour of a

larger metapopulation (e.g. Leberg 2005; Palstra & Ruzzante 2008) and provide a

harmonic mean of the Ne between the sampling points (Waples 2005). Because

very few of the oldest historical samples were collected prior to the major

demographic changes in the 19th century (see IV for details), and many

subsequent bottlenecks are likely to have occurred, the obtained Ne estimates

hardly represent that for a preceding undisturbed population. In fact, the Bayesian

method (Beaumont 1999; see also Girod et al. 2011) that was used to infer the

past demographic history in paper I strongly supported a long-term decline in the

Finnish wolf population. The contemporary population (1996–2004) was

estimated to be only ~8% of its historical size, and the ancient Ne (late 19th to

early 20th century) was thus about 590.

60

61

4 Conclusions

This study confirmed explicit patterns and rapid genetic changes in the

contemporary Finnish wolf population. Major findings included:

1. a low level of gene flow between the present-day Russian and Finnish wolf

populations and resulting genetic differentiation,

2. a relatively small local effective population size leading to gradual temporal

genetic changes in the population,

3. a significant increase in inbreeding during the recent heavy population

decline,

4. clear isolation by distance within the population due to the relatively short

mean dispersal distances of wolves and kin-based social structure and,

5. genetic deterioration of the contemporary Finnish wolf population compared

to the historical population, which was likely to be much larger and more

uniform with the Russian population.

On the basis of the results obtained in this study, it is clear that the Finnish wolf

population is currently too small (N = 120–135) to maintain a self-sufficient and

genetically healthy population in the long-term (e.g. Palstra & Ruzzante 2008).

The local effective population size should be considerably larger and/or the rate

of the effective immigration from the Russian population higher (i.e. the

immigrants should reach reproductive status in the Finnish wolf population) in

order to secure the population viability.

The question of population viability is not straightforward, however, and

exact numeric conservation goals are therefore difficult to provide. The ‘50/500’

rule of thumb (Franklin 1980) has often been used as a guiding principle in

conservation, and as such has become somewhat infamous and misused (e.g.

Flather et al. 2011; Jamieson & Allendorf 2012). The rationale behind the 50/500

rule is that the effective population size should be larger than 50 in order to avoid

extinction in the short term due to the harmful effects of inbreeding depression on

demography, whereas Ne ≥ 500 is necessary to retain the evolutionary potential of

a population (e.g. Jamieson & Allendorf 2012). In the management plan for the

wolf population in Finland (Anonymous 2005), it was stated that 20 breeding

pairs would secure the short-term viability, if the amount of gene flow from

Russian Karelia remains at the pre-existing level. On the other hand, over 25

breeding pairs would be needed if the level of immigration were to decrease.

Results from this study, showing a low degree of connectivity (I–III), support the

62

recommendation of at least 25 breeding pairs. So far this goal has only once been

reached, during a peak year in 2006 (Fig. 3).

The genetic viability concept is indisputably important and should be

integrated into the conservation of species (e.g. Frankham 2005; Laikre et al.

2009; Jamieson & Allendorf 2012). However, minimum conservation goals based

on effective size estimates alone are likely to be insufficient (see 1.3 and 1.3.1)

– not least due to the fact that Ne estimates themselves are easily biased and often

somewhat inaccurate (e.g. Luikart et al. 2010) and that the stability of the

Ne/Nc -ratio is not necessarily consistent (e.g. Palstra & Ruzzante 2008; III). For

example, based on the LD-Ne/Nc -ratios obtained in this study (III), ~180 wolves

would have been theoretically enough to avoid a significant increase of

inbreeding in the time period of 1998–2006 (Ne/Nc = 0.275–0.285), whereas as

many as 521 wolves would have been necessary to fulfil the minimum Ne

requirement of 50 during the last study period in 2007–2009 (Ne/Nc = 0.097).

By combining field studies and genetic analyses with the knowledge we have

on wolf population dynamics, it is possible to recognize the factors that lower the

effective population size, and thus adversely affect the population viability.

Because alpha wolves have very great significance in the maintenance of the

natural population structure and demography (e.g. Brainerd et al. 2008; Treves

2009), they should have the highest conservation priority. Additionally,

immigrants are crucial for the long-term survival of the population and thus their

recognition, for example using non-invasive genetic methods could be utilized in

conservation (e.g. Lucchini et al. 2002). From the management point of view, the

focus of genetic analyses should shift from retrospection to real-time monitoring.

The field data continuously collected by the Finnish Game and Fisheries Research

Institute and the genetic reference data-base collected in conjunction with this

study provide a good foundation for genetic monitoring in the future. However,

there is currently no ongoing funding for genetic monitoring.

This study provides diverse, new information on the current and historical

genetic status of the Finnish wolf population that can be utilized, for example, in

future wolf management. Ongoing genetic studies on wolves in our study group

by Alina Niskanen and Jenni Harmoinen concerning immune response-related

gene variation (AN) and the social-based population structure (JH) will

significantly increase and deepen our understanding of Finnish wolf genetics in

the near future. New molecular approaches, such as genome-wide SNP (single

nucleotide polymorphism) marker scans and high-throughput genome sequencing

63

technologies are also now readily available, and will most likely be adopted in

future studies concerning the Finnish wolf population.

64

65

References

Adams JR, Vucetich LM, Fredrick PW, Peterson RO & Vucetich JA (2011) Genomic sweep and potential genetic rescue during limiting environmental conditions in an isolated wolf population. Proceedings of the Royal Society B: Biological Sciences 278: 3336–3344.

Anonymous (2005) Management plan for the wolf population in Finland. Ministry of Agriculture and Forestry 11b/2005. URI: http://wwwb.mmm.fi/julkaisut/ julkaisusarja/2005/MMMjulkaisu2005_11b.pdf. Cited 27.08.2012.

Anonymous (1996) Suomen maasuurpetokannat ja niiden hoito. – Suurpetotyöryhmän raportti. Maa- ja Metsätalousministeriön julkaisuja 6/1996.

Allendorf FW & Luikart G (2007) Conservation and the genetics of populations. Blackwell Publishing.

Avise JC, Lansman RA & Shade RO (1979) The use of restriction endonucleases to measure mitochondrial DNA sequence relatedness in natural populations. I. Population structure and evolution in the genus Peromyscus. Genetics 92: 279–295.

Avise JC, Neigel JE & Arnold J (1984) Demographic influences of mitochondrial DNA lineage survivorship in animal populations. Journal of Molecular Evolution 20: 99–105.

Barker JSF (2011) Effective population size of natural populations of Drosophila buzzatii, with a comparative evaluation of nine methods of estimation. Molecular Ecology 20: 4452–4471.

Barnosky AD, Matzke N, Tomiya S, Wogan GOU, Swartz B, Quental E, Marshall C, McGuire C, Lindsey EL, Maguire KC, Mersey B & Ferrer EA (2011) Has the Earth’s sixth mass extinction already arrived? Nature 471: 51–57.

Barton NH & Slatkin M (1986) A Quasi-equilibrium theory of the distribution of rare alleles in a subdivided population. Heredity 56: 409–411.

Beaumont MA (1999) Detecting population expansion and decline using microsatellites. Genetics 153: 2013–2029.

Bisi J & Kurki S (2008) The wolf debate in Finland. Expectations and objectives for the management of the wolf population at regional and national level. Publications 12, Ruralia Institute, University of Helsinki, Seinäjoki. URI: http://www.helsinki.fi/ ruralia/julkaisut/pdf/Publications12.pdf. Cited 10.12.2012.

Bisi J, Kurki S, Svensberg M & Liukkonen T (2007) Human dimensions of wolf (Canis lupus) conflicts in Finland. European Journal of Wildlife Research 53: 304–314.

Bisi J, Liukkonen T, Mykrä S, Pohja-Mykrä M & Kurki S (2010) The good bad wolf – wolf evaluation reveals the roots of the Finnish wolf conflict. European Journal of Wildife research 56: 771–779.

Boitani L (2003) Wolf Conservation and Recovery. In Mech LD & Boitani L (eds) Wolves. Behaviour, Ecology and Conservation. Chicago, The University of Chicago Press: 317–340.

66

Boitani L (2000) Action plan for the conservation of the wolves (Canis lupus) in Europe. Convention on the conservation of European wildlife and natural habitats (Bern convention). Nature and environment, No. 113, Council of Europe Publishing.

Bouzat JL (2010) Conservation genetics of population bottlenecks: the role of chance, selection, and history. Conservation Genetics 11: 463–478.

Boyce MS (1992) Population viability analysis. Annual Review of Ecology and Systematics 23: 481–506.

Brainerd SM, Andrén H, Bangs EE, Bradley EH, Fontaine JA, Hall W, Iliopoulos Y, Jimenez MD, Jozwiak EA, Liberg O, Mach CM, Meier TJ, Niemeyer CC, Pedersen HC, Sand H, Schultz RN, Smith DW, Wabakken P & Wydeven AP (2008) The effects of breeder loss on wolves. Journal of Wildlife Management 72: 89–98.

Busch JD, Waser PM & DeWoody JA (2007) Recent demographic bottlenecks are not accompanied by a genetic signature in banner-tailed kangaroo rats (Dipodomys spectabilis). Molecular Ecology 16: 2450–2462.

Caniglia R, Fabbri E, Greco C, Galaverni M & Randi E (2010) Forensic DNA against wildlife poaching: Identification of a serial wolf killing in Italy. Forensic Science International: Genetics 4: 334–338.

Cardillo M, Mace GM, Jones KE, Bielby J, Bininda-Emonds ORP, Sechrest W, Orme CDL & Purvis A (2005) Multiple causes of high extinction risk in large mammal species. Science 309: 1239–1241.

Carmichael LE, Krizan J, Nagy A, Fuglei E, Dumond M, Johnson D, Veith A, Berteaux D & Strobek C (2007) Historical and ecological determinants of genetic structure in arctic canids. Molecular Ecology 16: 3466–3483.

Carmichael LE, Nagy JA, Larter NC & Strobek C (2001) Prey specialization may influence patterns of gene flow in wolves of the Canadian Northwest. Molecular Ecology 10: 2787–2798.

Casas-Marce M, Revilla E & Godoy JA (2010) Searching for DNA in museum specimens: a comparison of sources in mammal species. Molecular Ecology Resources 10: 502–507.

Ciucci P, Reggioni W, Maiorano L & Boitani L (2009) Long-distance dispersal of a rescued wolf from the Northern Apennines to the Western Alps. The Journal of Wildlife Management 73: 1300–1306.

Corander J, Waldmann P & Sillanpää MJ (2003) Bayesian analysis of genetic differentiation between populations. Genetics 163: 367–374.

Cornuet JM & Luikart G (1996) Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144: 2001–2014.

Croitor R & Brugal J-P (2010) Ecological and evolutionary dynamics of the carnivore community in Europe during the last 3 million years. Quaternary International 212: 98–108.

Croteau EK (2010) Causes and Consequences of Dispersal in Plants and Animals. Nature Education Knowledge 3(10): 12.

Danilov PI (2005) Game animals of Karelia. Ecology, resources, management, protection. Moscow, Nauka.

67

Dalerum F, Cameron EZ, Kunkel K & Somers MJ (2009) Diversity and depletions in continental carnivore guilds: implications for prioritizing global carnivore conservation. Biology Letters 5: 35–38.

Druzhkova AS, Thalmann O, Trifonov VA, Leonard JA, Vorobieva NV, Ovodov ND, Graphodatsky AS & Wayne RK (2013) Ancient DNA analysis affirms the Canid from Altai as a primitive dog. PLoS ONE 8(3): e57754

Edelaar P & Bolnick DI (2012) Non-random gene flow: an underappreciated force in evolution and ecology. Trends in Ecology and Evolution 27: 659–665.

Ellegren H (2004) Microsatellites: simple sequences with complex evolution. Nature Reviews Genetics 5: 435–445.

Ellegren H, Savolainen P & Rosén B (1996) The general history of an isolated population of the endangered grey wolf Canis lupus: a study of nuclear and mitochondrial polymorphism. Philosophical Transactions of the Royal Society, Series B 351: 1661–1669.

England PR, Luikart G & Waples RS (2010) Early detection of population fragmentation using linkage disequilibrium estimation of effective population size. Conservation Genetics 11: 2425–2430.

Enserink M & Vogel G (2006) Wildlife conservation—The carnivore comeback. Science, 374: 746–749.

Ericsson G & Heberlein TA (2003) Attitudes of hunters, locals, and the general public in Sweden now that the wolves are back. Biological Conservation 111: 149–159.

Ermala A (2003) A survey of large predators in Finland during the 19th-20th centuries. Acta Zoologica Lituanica 13: 15–20.

Estes JA, Terborgh J, Brashares JS, Power ME, Berger J, Bond WJ, Carpenter SR, Essington TE, Holt RD, Jackson JBC, Marquis RJ, Oksanen L, Oksanen T, Paine RT, Pikitch EK, Ripple WJ, Sandin SA, Scheffer M, Schoener TW, Shurin JB, Sinclair ARE, Soulé ME, Virtanen R & Wardle DA (2011) Trophic downgrading of planet Earth. Science 333: 301–306.

Excoffier L, Smouse P & Quattro J (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131: 479–491.

Fain SR, Straughan DJ & Taylor BF (2010) Genetic outcomes of wolf recovery in the western Great Lakes states. Conservation Genetics 11: 1747–1765.

Fernández JM & Ruiz de Azua N (2010) Historical dynamics of a declining wolf population: persecution vs. prey reduction. European Journal of Wildlife Research 56: 169–179.

Flagstad Ø, Walker CW, Vilá C, Sundqvist A-K, Fernholm B, Hufthammer AK, Wiig Ø, Kojola I & Ellegren H (2003) Two centuries of the Scandinavian wolf population: patterns of genetic variability and migration during an era of dramatic decline. Molecular Ecology 12: 869–880.

Flather CH, Hayward GD, Beissinger SR & Stephens PA (2011) Minimum viable populations: is there a ‘magic number’ for conservation practitioners? Trends in Ecology and Evolution 26: 307–316.

68

François O & Durand E (2010) Spatially explicit Bayesian clustering models in population genetics. Molecular Ecology Resources 10: 773–784.

Frankham R (2005) Genetics and extinction. Biological Conservation 126: 131–140. Frankham R (1996) Relationship of genetic variation to population size in wildlife.

Conservation Biology 10: 1500–1508. Frankham R (1995) Effective population size/ adult population size ratios in wildlife: e

review. Genetics Research 66: 95–107. Franklin I (1980) Evolutionary change in small populations. In Soulé ME & Wilcox BA

(eds) Conservation Biology: an Evolutionary–Ecological Perspective. Sinauer Associates: 135–150.

Frantz AC, Cellina S, Krier A, Schley L & Burke T (2009) Using spatial Bayesian methods to determine the genetic structure of a continuously distributed population: clusters or isolation by distance? Journal of Applied Ecology 46: 493–505.

Fuller T, Mech LD & Cochrane JF (2003) Wolf population dynamics. In Mech LD & Boitani L (eds) Wolves: Behavior, Ecology and Conservation. Chicago, The University of Chicago Press: 317–340.

Galtier N, Nabholz B, Glémin S & Hurst GDD (2009) Mitochondrial DNA as a marker of molecular diversity: a reappraisal. Molecular Ecology 18: 4541–4550.

Geffen E, Anderson MJ & Wayne RK (2004) Climate and habitat barriers to dispersal in the highly mobile grey wolf. Molecular Ecology 13: 2481–2490.

Geffen E, Kam M, Hefner R, Hersteinsson P, Angerbjörn A, Dalèn L, Fuglein E, Norèn K, Adams JR, Vucetih J, Meier TJ, Mech LD, vonHoldt BM, Stahler DR & Wayne RK (2011) Kin encounter rate and inbreeding avoidance in canids. Molecular Ecology 20: 5348–5358.

Gobush K, Kerr B & Wasser S (2009) Genetic relatedness and disrupted social structure in a poached population of African elephants. Molecular Ecology 18: 722–734.

Girod C, Vitalis R, Leblois R & Fréville H (2011) Inferring population decline and expansion from microsatellite data: a simulation-based evaluation of the Msvar method. Genetics 188: 165–179.

Grewal SK, Wilson PJ, Kung TK, Shami K, Theberge MT, Theberge JB & White BN (2004) A genetic assessment of the eastern wolf (Canis lycaon) in Algonquin Provincial Park. Journal of Mammalogy 85: 625–632.

Guillot G, Leblois R, Coulon A & Frantz AC (2009) Statistical methods in spatial genetics. Molecular Ecology 18: 4734–4756.

Haag T, Santos AS, Sana DA, Morato RG, Cullen L Jr, Crawshaw PG Jr, De Angelo C, Di Bitetti MS, Salzano FM & Eizirik E (2010) The effect of habitat fragmentation on the genetic structure of a top predator: loss of diversity and high differentiation among remnant populations of Atlantic Forest jaguars (Panthera onca). Molecular Ecology 19: 4906–4921.

Hansen MM, Andersen LW, Aspi J & Fredrickson R (2011) Evaluation of the conservation genetic basis of management of grey wolves in Sweden. URI: http://www.artdata.slu.se/filer/Evaluation-of-the-conservation-genetic-basis-of-management-of-grey-wolves-in-Sweden.pdf. Cited 21.1.2013.

69

Hardy OJ & Vekemans X (1999) Isolation by distance in a continuous population: reconciliation between spatial autocorrelation analysis and population genetic models. Heredity 83: 145–154.

Harris H (1966) Enzyme polymorphism in man. Proceedings in Royal Society of London, B 164: 298–310.

Hebblewhite M, White CA, Nietvelt CG, McKenzie JA, Hurd TE, Fryxell JM, Bayley SE & Paquet PC (2005) Human activity mediates a trophic cascade caused by wolves. Ecology 86: 2135–2144.

Hedrick PW & Kalinowski ST (2000) Inbreeding depression in Conservation Biology. Annual Review of Ecology and Systematics. 31: 139–162.

Hewitt G (2000) The genetic legacy of the Quaternary ice ages. Nature 405: 907–913. Hindrikson M, Männil P, Ozolins J, Krzywinski A & Saarma U (2012) Bucking the trend

in wolf-dog hybridization: First evidence from Europe of hybridization between female dogs and male wolves. PLoS ONE 7(10): e46465.

Hoehn M, Gruber B, Sarre SD, Lange R & Henle K (2012) Can genetic estimators provide robust estimates of the effective number of breeders in small populations? PLoS ONE 7(11): e48464.

Hoelzel AR, Hey J, Dahlheim ME, Nicholson C, Burkanov V & Black N (2007) Evolution of population structure in a highly social top predator, the Killer whale. Molecular Biology and Evolution 24: 1407–1415.

Holsinger KE & Weir BS (2009) Genetics in geographically structures populations: defining, estimating and interpreting FST. Nature Reviews Genetics 10: 639–650.

Jamieson IG & Allendorf FW (2012) How does the 50/500 rule apply to MVPs? Trends in Ecology and Evolution 27(10): 578–584.

Jędrzejewski W, Branicki W, Veit C, Medugorac I, Pilot M, Bunevich AN, Jędrzejewska B, Schmidt K, Theuerkauf J, Okarma H, Gula R, Szymura L & Förster M (2005) Genetic diversity and relatedness within packs in an intensely hunted population of wolves Canis lupus. Acta Theriologica 50: 3–22.

Kaartinen S, Luoto M & Kojola I (2010) Selection of den sites by wolves in boreal forests in Finland. Journal of Zoology 281: 99–104.

Kaartinen S, Kojola I & Colpaert A (2005) Finnish wolves avoid roads and settlements. Annales Zoologici Fennici 42: 523–532.

Kalinowski ST (2004) Counting alleles with rarefaction: Private alleles and hierarchical sampling designs. Conservation Genetics, 5: 539–543.

Keller LF & Waller DM (2002) Inbreeding effects in wild populations. Trends in Ecology and Evolution 17: 230–241.

Kennedy PK, Kennedy ML, Clarkson PL & Liepins IS (1991) Genetic variability in natural populations of the gray wolf, Canis lupus. Canadian Journal of Zoology 69: 1183–1188.

Kiiskilä J (2006) Suden (Canis lupus) geneettinen rakenne Suomessa ja Euraasian susien fylogeografia. University of Oulu, Finland. (Master’s thesis in Finnish)

70

Kim KS, Lee SE, Jeong HW & Ha JH (1998) The complete nucleotide sequence of the domestic dog (Canis familiaris) mitochondrial genome. Molecular Phylogenetics and Evolution 10: 210–220.

Kojola I, Helle P & Heikkinen S (2011) Susikannan viimeaikaiset muutokset Suomessa eri aineistojen valossa. Suomen Riista 57: 55–62.

Kojola I, Aspi J, Hakala A, Heikkinen S, Ilmoni C & Ronkainen S (2006) Dispersal in an expanding wolf population in Finland. Journal of Mammalogy 87: 281–286.

Kojola I & Määttä E (2004) Suurpetojen lukumäärä ja lisääntyminen vuonna 2003. Riistantutkimuksen Tiedote 194: 1–7.

Kopatz A, Eiken HG, Hagen SB, Ruokonen M, Esparza-Salas R, Schregel J, Kojola I, Smith ME, Wartiainen I, Aspholm PE, Wikan S, Rykow AM, Makarova O, Polikarpova N, Tirronen KF, Danilov PI & Aspi J (2012) Connectivity and population subdivision at the fringe of a large brown bear (Ursus arctos) population in North Western Europe. Conservation Genetics 13: 681–692.

Laikre L, Nilsson T, Primmer CR, Ryman N & Allendorf FW (2009) Importance of genetics in the interpretation of favourable conservation status. Conservation Biology 23: 1378–1381.

Laikre L, Ryman N & Thompson EA (1993) Hereditary Blindness in a Captive Wolf (Canis lupus) Population: Frequency Reduction of a Deleterious Allele in Relation to Gene Conservation. Conservation Biology 7: 592–601.

Lehman N, Clarkson P, Mech LD, Meier TJ & Wayne RK (1992) The use of DNA fingerprinting and mitochondrial DNA to study genetic relationships within and among wolf packs. Behavioral Ecology and Sociobiology 30: 83–94.

Lewontin RC & Hubby JL (1966) A molecular approach to the study of genetic heterozygosity in natural populations. II. Amount of variation and degree of heterozygosity in natural populations of Drosophila pseudoobscura. Genetics 54: 595–605.

Li W-C (1978) Maintenance of genetic variability under the joint effect of mutation, selection and random drift. Genetics 90: 349–382.

Li Y-C, Korol AB, Fahima T, Beiles A & Nevo E (2002) Microsatellites: genomic distribution, putative functions and mutational mechanisms: a review. Molecular Ecology 11: 2453–2465.

Liberg O, Andrén H, Pedersen H-C, Sand H, Sejberg D, Wabakken P, Åkesson M & Bensch S (2005) Severe inbreeding depression in a wild wolf (Canis lupus) population. Biology Letters 1: 17–20.

Liberg O, Chapron G, Wabakken P, Pedersen HC, Hobbs NT & Sand H (2012) Shoot, shovel and shut up: cryptic poaching slows restoration of a large carnivore in Europe. Proceedings of the Royal Society B 279: 910–915.

Linnell J, Salvatori V & Boitani L (2008). Guidelines for population level management plans for large carnivores in Europe. A Large Carnivore Initiative for Europe report prepared for the European Commission (contract 070501/2005/424162/MAR/B2).

Lowe WH & Allendorf FW (2010) What can genetics tell us about population connectivity? Molecular Ecology 19: 3038–3051.

71

Lucchini V, Galov A & Randi E (2004) Evidence of genetic distinction and long-term population decline in wolves (Canis lupus) in the Italian Apennines. Molecular Ecology 13: 523–536.

Lucchini V, Fabri E, Marucco F, Ricci S, Boitani L & Randi E (2002) Noninvasive molecular tracking of colonizing wolf (Canis lupus) in the western Italian Alps. Molecular Ecology 11: 857–868.

Luikart G, Ryman N, Tallmon DA, Schwartz MK & Allendorf FA (2010) Estimation of census and effective population sizes: the increasing usefulness of DNA-based approaches. Conservation Genetics 11: 355–373.

Marsden CD, Woodroffe R, Mills MGL, McNutt JW, Creel S, Groom R, Emmanuel M, Cleaveland S, Kat P, Rasmussen GSA, Ginsberg J, Lines R, André J-M, Begg C, Wayne RK & Mable BK (2012) Spatial and temporal patterns of neutral and adaptive genetic variation in the endangered African wild dog (Lycaon pictus). Molecular Ecology 21: 1379–1393.

Martínková N & Searle JB (2006) Amplification success rate of DNA from museum skin collections: a case study of stoats from 18 museums. Molecular Ecology Notes 6: 1014–1017.

Mech LD (2012) Is science in danger of sanctifying the wolf? Biological Conservation 150: 143–149.

Mech LD (1974) Canis lupus. Mammalian Species 37: 1–6. Mech LD & Boitani L (2004) 5.2. Grey wolf Canis lupus. In Sillero-Zubiri C, Hoffmann

M & Macdonald DW (eds) Canids: Foxes, Wolves, Jackals and Dogs. Status Survey and Conservation Action Plan. IUCN/SSC Canid Specialist Group. Gland, Switzerland and Cambridge, UK: 124–129.

Mech LD & Boitani L (2003) Wolves social ecology. In Mech LD & Boitani L (eds) Wolves: Behavior, ecology and conservation. Chicago, The University of Chicago Press: 1–34.

Meirmans PG (2012) The trouble with isolation by distance. Molecular Ecology 21: 2839–2846.

Muñoz-Fuentes V, Darimont CT, Wayne RK, Paquet P & Leonard JA (2009) Ecological factors drive differentiation in British Columbia grey wolves. Journal of Biogeography 36: 1516–1531.

Musiani M, Leonard JA, Cluff HD, Gates CC, Mariani S, Paqueat PC, Vilà C & Wayne RK (2007) Differentiation of tundra/taiga and boreal coniferous forest wolves: genetics, coat colour and association with migratory caribou. Molecular Ecology 16: 4149–4170.

Mäensyrjä P (1974) Hukka huutaa. Hämeenlinna, Karisto. Nunney L & Elam DR (1994) Estimating the effective size of conserved populations.

Conservation Biology 8: 175–184. Nyström V, Angerbjörn A & Dalén L (2006) Genetic consequences of a demographic

bottleneck in the Scandinavian arctic fox. Oikos 114: 84–94.

72

O’Grady JJ, Brook BW, Reed DH, Ballou JD, Tonkyn DW & Frankham R (2006) Realistic levels of inbreeding depression strongly affect extinction risk in wild populations. Biological Conservation 133: 42–51.

Palstra FP & Ruzzante DE (2008) Genetic estimates of contemporary effective population size: what can they tell us about the importance of genetic stochasticity for wild population persistence? Molecular Ecology 17: 3428–3447.

Pang J-F, Kluetsch C, Zou X-J, Zhang A-b, Luo L-Y, Angleby H, Ardalan A, Ekström C, Sköllermo A, Lundebeg J, Matsumara S, Leitner T, Zhang Y-P & Savolainen P (2009) mtDNA data indicate a single origin for dogs south of Yangtze river, less than 16,300 years ago, from numerous wolves. Molecular Biology and Evolution 26: 2849–2864.

Peery MZ, Kirby R, Reid BN, Stoelting R, Doucet-Bëer E, Robinson S, Vásquez-Carrillo C, Pauli JN & Palsbøll PJ (2012) Reliability of genetic bottleneck tests for detecting recent population declines. Molecular Ecology 21: 3403–3418.

Pilot M, Jȩdrzejewski W, Sidorovich VE, Meier-Augenstein W & Hoelzel R (2012) Dietary differentiation and evolution of population genetic structure in a highly mobile carnivore. PLoS ONE 7(6): e39341.

Pilot M, Branicki W, Jȩdrzejewski W, Goszcyńki J, Jȩdrzejevska B, Dykyy I, Shkvyrya M & Tsingarska E (2010) Phylogeographic history of grey wolves in Europe. Evolutionary Biology 10: 104.

Pilot M, Jędrzejewski W, Branicki W, Sidorovich VE, Jędrzejevska B, Stachura K & Funk SM (2006) Ecological factors influence population genetic structure of European grey wolves. Molecular Ecology 15: 4533–4553.

Pope LC, Domingo-Roura X, Erven K & Burke T (2006) Isolation by distance and gene flow in the Eurasian badger (Meles meles) at both a local and broad scale. Molecular Ecology 15: 371–386.

Pritchard JK, Stephens M & Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155: 945–959.

Pulliainen E (2011) Suurpetoja tarkkailtu rajoilla vuosikymmeniä. Rajamme Vartijat 2/2011: 54.

Pulliainen E (1980) The status, structure and behaviour of the wolf (Canis lupus L.) along the Fenno-Soviet border. Annales Zoologici Fennici 17: 107–112.

Pulliainen E (1965) Studies on the wolves (Canis lupus L.) in Finland. Annales Zoologici Fennici 2: 215–259.

Ramakrishnan U & Hadly EA (2009) Using phylochronology to reveal cryptic population histories: review and synthesis of 29 ancient DNA studies. Molecular Ecology 18: 1310–1330.

Randi E (2011) Genetics and conservation of wolves Canis lupus in Europe. Mammal Review 41: 99–111.

Randi E, Lucchini V, Christensen MF, Mucci N, Funk SM, Dolf G & Loeschcke V (2000) Mitochondrial DNA variability in Italian and East European wolves: Detecting the consequences of small population size and hybridization. Conservation Biology 14: 464–473.

73

Rannala B & Mountain JL (1997) Detecting immigration by using multilocus genotypes. PNAS 94: 9197–9221.

Rassi P, Hyvärinen E, Juslén A & Mannerkoski I (eds) (2010) The 2010 Red List of Finnish Species. Ministry of the Environment and Finnish Environment Institute.

Reed DH, O’Grady JJ, Brook BW, Ballou JD & Frankham R (2003) Estimates of minimum viable population sizes for vertebrates and factors influencing those estimates. Biological Conservation 113: 23–34.

Rieseberg LH & Burke JM (2001) A genic view of species integration. Journal of Evolutionary Biology 14: 883–886.

Ripple WJ & Beschta RL (2012) Trophic cascades in Yellowstone: The first 15 years after wolf reintroduction. Biological Conservation 145: 205–213.

Rodríguez-Ramilo ST & Wang J (2012) The effect of close relatives on unsupervised Bayesian clustering algorithms in population genetic structure analysis. Molecular Ecology Resources 12: 873–884.

Rohland N & Hofreiter M (2007) Ancient DNA extraction from bones and teeth. Nature protocols 2(7): 1756–1762.

Roy MS, Geffen E, Smith D, Ostrander EA & Wayne RK (1994) Patterns of differentiation and hybridization in North American wolflike canids, revealed by analysis of microsatellite loci. Molecular Biology and Evolution 11(4): 553–570.

Rutledge LY, White BN, Row JR & Patterson BR (2011) Intense harvesting of eastern wolves facilitated hybridization with coyotes. Ecology and Evolution 2: 19–33.

Rutledge LY, Patterson BR, Mills KJ, Loveless KM, Murray DL & White BN (2010) Protection from harvesting restores the natural social structure of eastern wolf packs. Biological Conservation 143: 332–339.

Räikkönen J, Bignert A, Mortensen P & Fernholm B (2006) Congenital defects in a highly inbred wild wolf population (Canis lupus). Mammalian Biology 71: 65–73.

Räikkönen J, Vucetih JA, Peterson RO & Nelson MP (2009) Congenital bone deformities and the inbred wolves (Canis lupus) of Isle Royale. Biological Conservation 142: 1025–1031.

Sastre N, Vilà C, Salinas M, Bologov VV, Urios V, Sánches A, Francino O & Ramírez O (2011) Signatures of demographic bottlenecks in European wolf populations. Conservation Genetics 12: 701–712.

Scandura M, Iacolina L, Capitani C, Gazzola A, Mattioli L & Apollonio M (2011) Fine-scale genetic structure suggests low levels of short-range gene flow in a wolf population of the Italian Apennines. European Journal of Wildlife Research 57: 949–958.

Schlötterer C (2004) The evolution of molecular markers – just a matter of fashion? Nature Reviews Genetics 5: 63–69.

Seddon JM, Sundqvist A-K, Björnerfeldt S & Ellegren H (2006) Genetic identification of immigrants to the Scandinavian wolf population. Conservation Genetics 7: 225–230.

Selander RK & Yang SY (1969) Protein polymorphism and genic heterozygosity in a wild population of the house mouse (Mus musculus). Genetics 63: 653–667.

74

Sergio F, Caro T, Brown D, Clucas B, Hunter J, Ketchum J, McHugh K & Hiraldo F (2008) Top predators as conservation tools: ecological rationale, assumptions, and efficacy. Annual Reviews of Ecology, Evolution and Systematics 39: 1–19.

Slatkin M (1985) Rare alleles as indicators of gene flow. Evolution 39: 53–65. Smith DW, Meier T, Geffen E, Mech LD, Burch JW, Adams LG & Wayne RK (1997) Is

incest common in gay wolf packs? Behavioral Ecology 8: 384–391. Smith DW, Peterson RO & Houston DB (2003) Yellowstone after wolves. Bioscience 53:

330–340. Sotelo G, Morán P, Fernández L & Posada D (2008) Genetic variation of the spiny spider

crab Maja brachydactyla in the northwestern Atlantic. Marine Ecology Progress series 362: 211–223.

Stenglein JL, Waits LP, Ausband DE, Zager P & Mack CM (2011) Estimating gray wolf pack size and family relationships using noninvasive genetic sampling at rendezvous sites. Journal of Mammalogy 92: 784–795.

Stronen AV, Forbes GJ, Paquet PC, Goulet G, Sallows T & Musiani M (2012) Dispersal in a plain landscape: short-distance genetic differentiation in southwestern Manitoba wolves, Canada. Conservation Genetics 13: 359–371.

Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585–595.

Tallmon DA, Koyuk A, Luikart G & Beaumont MA (2008) ONeSAMP: a program to estimate effective population size using approximate Bayesian computation. Molecular Ecology Resources 8: 299–301.

Thalmann O, Wegmann D, Spitzner M, Arandjelovic M, Guschanski K, Leuenberger C, Bergl RA & Vigilant L (2011) Historical sampling reveals dramatic demographic changes in western gorilla populations. BMC Evolutionary Biology 11: 85.

Taylor CM & Norris DR (2007) Predicting conditions for migration: effects of density dependence and habitat quality. Biology Letters 3: 280–283.

Traill LW, Brook BW, Frankham RR & Bradshaw CJA (2010) Pragmatic population viability targets in a rapidly changing world. Biological Conservation 143: 28–34.

Traill LW, Bradshaw JA & Brook BW (2007) Minimum viable population size: A meta-analysis of 30 years of published estimates. Biological Conservation 139: 159–166.

Treves A (2009) Hunting for large carnivore conservation. Journal of Applied Ecology 46: 1350–1356.

Ugelvig LV, Nielsen PS, Boomsma JJ & Nash DR (2011) Reconstructing eight decades of genetic variation in an isolated Danish population of the large blue butterfly Maculinea arion. BMC Evolutionary Biology 11: 201.

Valière N, Fumagalli L, Gielly L, Miquel C, Lequette B, Poulle M-L, Weber J-M, Arlettaz R & Taberlet P (2003) Long-distance wolf recolonization of France and Switzerland inferred from non-invasive genetic sampling over a period of 10 years. Animal Conservation 6: 83–92.

Vekemans X, Hardy OJ (2004) New insights from fine-scale spatial genetic structure analysis in plant populations. Molecular Ecology 13: 921–935.

75

Verardi A, Lucchini V & Randi E (2006) Detecting introgressive hybridization between free-ranging domestic dogs and wild wolves (Canis lupus) by admixture linkage disequilibrium analysis. Molecular Ecology 15: 2845–2855.

Vilà C, Amorim IR, Leonard JA, Posada D, Castroviejo J, Petrucci-Fonseca F, Crandall KA, Ellegren H & Wayne RK (1999) Mitochondrial DNA phylogeography and population history of the grey wolf Canis lupus. Molecular Ecology 8: 2089–2103.

Vilà C, Sundqvist A-K, Flagstad Ø, Seddon J, Björnerfeldt S, Kojola I, Casulli A, Sand H, Wabakken P & Ellegren H (2003a) Rescue of a severely bottlenecked wolf (Canis lupus) population by a single immigrant. Proceedings of the Royal Society of London. Series B, Biological Sciences 270: 91–97.

Vilà C, Walker C, Sundqvist AK, Flagstad Ø, Seddon J, Björnefeldt S, Kojola I, Casulli A, Sand H, Wabakken P & Ellegren H (2003b) Combined use of maternal, paternal and bi-parental genetic markers for the identification of wolf-dog hybrids. Heredity 90: 17–24.

vonHoldt B, Pollinger JP, Earl DA, Knowles JC, Boyko AR, Parker H, Geffen E, Pilot M, Jedrzejewski W, Jedrzejewska B, Sidorovich V, Greco C, Randi E, Musiani M, Kays R, Bustamante CD, Ostrander EA, Novembre J & Wayne RK (2011) A genome-wide perspective on the evolutionary history of enigmatic wolf-like canids. Genome Research 21: 1294–1305.

vonHoldt BM, Stahler D, Bangs EE, Smith DW, Jimenez MD, Mack CM, Niemeyer CC, Pollinger JP & Wayne RK (2010) A novel assessment of population structure and gene flow in grey wolf populations of the Northern Rocky Mountains of the United States. Molecular Ecology 19: 4412–4427.

vonHoldt BM, Stahler DR, Smith DW, Earl DA, Pollinger JP & Wayne (2008) The genealogy and genetic viability of reintroduced Yellowstone grey wolves. Molecular Ecology 17: 252–274.

Wabakken P, Sand H, Liberg O & Bjärvall A (2001) The recovery, distribution and population dynamics of wolves on the Scandinavian Peninsula. Canadian Journal of Zoology 79: 710–725.

Wandeler P, Hoeck PEA & Keller LK (2007) Back to the future: museum specimens in population genetics. Trends in Ecology and Evolution 22: 634–642.

Wang J (2005) Estimation of effective population size from temporally spaced samples. Philosophical Transactions of the Royal Society B 360: 1395–1409.

Wang J & Whitlock MC (2003) Estimating effective population size and migration rates from genetic samples over space and time. Genetics 163: 429–446.

Waples RS (2010) Spatial-temporal stratifications in natural populations and how they affect understanding and estimation of effective population size. Molecular Ecology Resources 10: 785–796.

Waples RS (2005) Genetic estimates of contemporary effective population size: to what time periods do the estimates apply? Molecular Ecology 14: 3335–3352.

Waples RS (2002) Definition and estimation of effective population size in the conservation of endangered species. In Beissinger SR & McCullough DR (eds) Population Viability Analysis. Chicago, University of Chicago Press: 147–168.

76

Waples RS & Do C (2008) LDNe: a program for estimating effective population size based on linkage disequilibrium. Molecular Ecology Resources 8: 753–756.

Waples RS & Gaggiotti O (2006) What is a population? An empirical evaluation of some genetic methods for identifying the number of gene pools and their degree of connectivity. Molecular Ecology 15: 1419–1439.

Wayne RK, Geffen E & Vilà C (2004) Conservation genetics of canids. In Sillero-Zubiri C, Hoffmann M & Macdonald DW (eds) Canids: Foxes, Wolves, Jackals and Dogs. Status Survey and Conservation Action Plan. Gland Switzerland and Cambridge UK, IUCN/SSC Canid Specialist Group: 237–245.

Wayne RK & Morin PA (2004) Conservation genetics in the new molecular age. Frontiers in Ecology and the Environment 2: 89–97.

Wayne RK, Lehman N, Allard MW & Honeycutt RL (1991) Mitochodrial DNA variability of the gray wolf: genetic consequences of population decline and habitat fragmentation. Conservation Biology 6: 559–569.

Weckworth BV, Talbot S, Sage GK, Person DK & Cook J (2005) A signal for independent coastal and continental histories among North American wolves. Molecular Ecology 14: 917–930.

Whitlock MC & McCauley DE (1999) Indirect measures of gene flow and migration: FST ≠ 1/(4Nm + 1). Heredity 82: 117–125.

Willi Y, Van Buskirk J & Hoffmann AA (2006) Limits to the adaptive potential of small populations. Annual Reviews of Ecology, Evolution and Systematics 37: 433–458.

Williamson-Natesan EG (2005) Comparison of methods detecting bottlenecks from microsatellite loci. Conservation Genetics 6: 551–562.

Wilson GA & Rannala B (2003) Bayesian inference of recent migration rates using multilocus genotypes. Genetics 163: 1177–1191.

Woodroffe R (2000) Predators and people: using human densities to interpret declines in large carnivores. Animal Conservation 3: 165–173.

Wright S (1951) The genetical structure of natural populations. Annals of Eugenics 15: 323–354.

Wright S (1946) Isolation by distance under diverse systems of mating. Genetics 31: 39–59. Wright (1943) Isolation by distance. Genetics 28: 114–138. Wright S (1931) Evolution in Mendelian populations. Genetics 16: 97–159. Yang J & Jiang Z (2011) Genetic diversity, population genetic structure and demographic

history of Przewalski’s gazelle (Procapra przewalskii): implications for conservation. Conservation Genetics 12: 1457–1468.

Zimmermann J, Hajibabaei M, Blckburn DC, Hanken J, Cantin E, Posfai J & Evans TC Jr (2008) DNA damage in preserved specimens and tissue samples: a molecular assessment. Frontiers in Zoology 5: 18.

77

Original articles

I Aspi J, Roininen E, Ruokonen M, Kojola I & Vilà C (2006) Genetic diversity, population structure, effective population size and demographic history of the Finnish wolf population. Molecular Ecology 15: 1561–1576.

II Aspi J, Roininen E, Kiiskilä J, Ruokonen M, Kojola I, Bljudnik L, Danilov P, Heikkinen S & Pulliainen E (2009) Genetic structure of the northwestern Russian wolf populations and gene flow between Russia and Finland. Conservation Genetics 10: 815–826.

III Jansson E, Ruokonen M, Kojola I & Aspi J (2012) Rise and fall of a wolf population: Genetic diversity and structure during recovery, rapid expansion, and drastic decline. Molecular Ecology 21: 5178–5193.

IV Jansson E, Harmoinen J, Ruokonen M & Aspi J (2013) Reconstructing the genetic history of the Finnish wolf population during the last 150 years. Manuscript.

Reprinted with permission from John Wiley and Sons (I, III) and Springer (II).

Original publications are not included in the electronic version of the dissertation.

78

A C T A U N I V E R S I T A T I S O U L U E N S I S

Book orders:Granum: Virtual book storehttp://granum.uta.fi/granum/

S E R I E S A S C I E N T I A E R E R U M N A T U R A L I U M

593. Tiikkaja, Marjo (2012) Value creation in collaboration between software suppliersand customers: suppliers’ perspective

594. Rousu, Timo (2012) Liquid chromatography–mass spectrometry in drugmetabolism studies

595. Kangas, Teija (2012) Theoretical study of the oxidation of a pure and alloyedcopper surface

596. Härkönen, Laura (2012) Seasonal variation in the life histories of a viviparousectoparasite, the deer ked

597. Niinimäki, Sirpa (2012) Reconstructing physical activity from human skeletalremains : potentials and restrictions in the use of musculoskeletal stress markers

598. Mandić, Vladimir (2012) Measurement-based value alignment and reasoning aboutorganizational goals and strategies : Studies with the ICT industry

599. Leiviskä, Katja (2012) Why information systems and software engineeringstudents enter and leave their study programme : A factor model and processtheory

600. Siira, Tuula (2012) Value Creation by Enterprise Systems Value Added Resellers :The Case of PLM Systems VARs

601. Kontula, Jukka (2012) New venture creation in software business : A contextuallyembedded entrepreneur’s perspective

602. Juntunen, Kaisu (2012) Tieto- ja viestintätekniikan soveltamiseen perustuvattoimintaprosessien uudistukset terveydenhuollossa : Sosio-teknis-taloudellinennäkökulma

603. Seppä, Karri (2012) Quantifying regional variation in the survival of cancerpatients

604. Kuvaja, Pasi (2012) Software process capability and maturity determination :BOOTSTRAP methodology and its evolution

605. Laanti, Maarit (2012) Agile Methods in large-scale software developmentorganizations : applicability and model for adoption

606. Hokkanen, Juho (2013) Liquid chromatography/mass spectrometry of bioactivesecondary metabolites – in vivo and in vitro studies

607. Kuokkanen, Matti (2013) Development of an eco- and material-efficient pelletproduction chain—a chemical study

ABCDEFG

UNIVERS ITY OF OULU P.O.B . 7500 F I -90014 UNIVERS ITY OF OULU F INLAND

A C T A U N I V E R S I T A T I S O U L U E N S I S

S E R I E S E D I T O R S

SCIENTIAE RERUM NATURALIUM

HUMANIORA

TECHNICA

MEDICA

SCIENTIAE RERUM SOCIALIUM

SCRIPTA ACADEMICA

OECONOMICA

EDITOR IN CHIEF

PUBLICATIONS EDITOR

Senior Assistant Jorma Arhippainen

University Lecturer Santeri Palviainen

Docent Hannu Heusala

Professor Olli Vuolteenaho

University Lecturer Hannu Heikkinen

Director Sinikka Eskelinen

Professor Jari Juga

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-0115-3 (Paperback)ISBN 978-952-62-0116-0 (PDF)ISSN 0355-3191 (Print)ISSN 1796-220X (Online)

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

OULU 2013

A 608

Eeva Jansson

PAST AND PRESENT GENETIC DIVERSITY AND STRUCTURE OF THE FINNISH WOLF POPULATION

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF SCIENCE,DEPARTMENT OF BIOLOGY

A 608

ACTA

Eeva Jansson